Upward Influence Tactics and their Effects on Job Performance Ratings and Flexible Working Arrangements: The Mediating Roles of Mutual Recognition Respect and Mutual Appraisal Respect

Supervisor’s HR decisions have a significant impact on the employees they manage but have been found to be subject to bias. The upward influence tactics use by subordinates can play a role in this. We investigated the effects of seven upward influence tactics on supervisor job performance ratings and the extent of subordinate flexible working arrangements (FWAs). Supervisors are often responsible for determining whether employees are granted FWAs. We posit an alternative theoretical mechanism by which upward influence tactics bring about their effects, mediated through two distinct types of respect, mutual appraisal and mutual recognition respect. We collected data from 389 matched supervisor-subordinate dyads, and found that both mutual appraisal respect and mutual recognition respect mediated relationships between several upward influence tactics and both job performance ratings and flexible working arrangements. Our findings show that upward influence tactics affect the quality of the relationship between employees and their supervisors. Specifically, these two forms of mutual respect. Further, that both mutual appraisal and mutual recognition respect may explain why supervisors show favoritism to some subordinates over others, in reaching HR decisions in these areas.

One of the most significant transformations in the field of human resource management over the past few decades has been the increasing responsibility that line managers now have in making human resource decisions (Brewster, Brookes, & Gollan, 2015;Purcell & Hutchinson, 2007). Indeed, how line managers interpret HR policies and enact decisions is seen as a critical factor explaining why the results of HR strategies are often variable (Bowen & Ostroff, 2004;Clarke, 2006;Williams, McDonald & Cathcart, 2016). Yet, employees can also play an active role in influencing key decisions in which they perceive they have a greater personal stake. Chief among these, are decisions line managers make in relation to performance ratings, promotion, career development, as well as salary increases. Since Kipnis, Schmidt, and Wilkinson´s early work in 1980, a significant body of literature has accumulated showing that employees´ use of particular influence tactics (called upward influence tactics) is associated with line manager decisions in these areas (Lee et al., 2017;Terpstra-Tong & Ralston, 2002). Upward influence tactics have been defined as the behavior used to gain compliance or obtain a desired goal from those at higher levels in the organization (Kipnis, Schmidt, & Wilkinson, 1980;Farmer, et al., 1997). Research has shown that positive benefits can accrue to employees as a result of their use of influence tactics. For example, significant relationships have been found between a range of upward influence tactics and job performance ratings (Higgins, Judge, & Ferris, 2003;Yukl & Tracey, 1992), assessment of promotability (Thacker & Wayne, 1995), salary increases (Kipnis & Schmidt, 1988), and enhanced trust from their managers (Su, 2010). There is now a major body of research showing these influence tactics to differentially affect a range of outcomes. For example, rational persuasion and ingratiation have been found to be positively related to supervisor job performance ratings and promotability, whilst assertiveness and coalition have Influence Tactics, Respect and Human Resource Decisions 3 generally been found to be negatively related (Kipnis & Schmidt, 1988;Higgins et al., 2003;Rao, Schmidt, & Murray, 1995;Su, 2010;Thacker & Wayne, 1995;Wayne et al., 1997). Others have also pointed to the use of particular upward influence tactics as having wider organizational consequences. This has included their effects on organizational culture, communication and decision-making processes, as well as the quality of workplace relationships more generally (Ansari, Aafaqi, & Zainal, 2007;Cable & Judge, 2003;Ralston et al., 1993).
For the most part, research in the area of upward influence tactics has focused on factors that affect the choice and effects of these tactics on human resource decisions.
Consequently, a range of individual (both agent and target) as well situational factors have been identified in the literature. Individual factors include belief systems, personality, and motivations (Rao et al., 1995). Whilst situational factors include aspects such as management style, relationship quality, and spatial distance (Cable & Judge 2003;Ferris & Judge 1991;Farmer et al., 1997). Other studies have examined how cultural differences affect both the choice and effects of influence tactics (Fu et al., 2004;Qiadan, Tziner & Waismel-Manor, 2012;Wang & Baiyin, 2017). Whilst this body of work has contributed to a better understanding of employee preferences for particular upward influence tactics over others, far less research has sought to explain how the use of upward influence tactics affect supervisor HR decisions.
One of the chief explanations posited explaining these effects is through social information processing (Hastie & Park, 1986;Lord & Harvey, 2002). The use of these tactics by employees is thought to influence supervisor attributions of their behavior.
These are then encoded into overall judgements as whether the employee is either likeable/dislikeable or competent/incompetent. These then become activated in future situations, including when making human resource decisions that affect the employee  (Srull & Wyer 1989;Wayne & Ferris 1990). Some tactics create goodwill, while others can create threat or pressure that affect the recipient's liking for the subordinate.
Importantly, liking has been found to be positively related to supervisor reward behavior (Ferris, et al., 1994). Others have also suggested that an employee's use of influence tactics will affect employee and supervisor appraisals of the quality of their relationship more broadly (Dienesch & Liden, 1986;Schriesheim, Castro & Yammarino, 2000). This draws upon social exchange theory (Blau, 1964), in positing that supervisors will respond more favorably in their HR decisions to those employees with whom they have relationships they value. Consequently, relationship quality also has been posited as a possible mediator between upward influence tactics and human resource decisions (Nahrgang & Seo, 2015). Specifically in relation to job performance ratings, this corresponds with the notion that social context significantly influences the performance appraisal process (Ferris et al., 2008;Levy & Williams, 2004;Pichler, 2012).
The first objective of our study is to build on this perspective in positing mutual respect as a mediating mechanism, by which upward influence tactics might affect supervisor HR decisions. We suggest that employees' use of particular influence tactics will positively or negatively affect the mutual respect they have for one another.
Furthermore, that higher levels of mutual respect will in turn, lead to more positive HR decision outcomes. We extend our theorizing by incorporating both mutual appraisal respect and mutual recognition respect as two distinct forms of the respect construct.
Mutual appraisal respect is a form of respect that reflects an individual's status or standing.
Mutual recognition respect (sometimes referred to as equality-based respect) refers to the respect we are due based on being treated fairly (Clarke, 2011). Our second objective is to investigate the effects of upward influence tactics on employees' flexible work arrangements (FWAs) in addition to job performance ratings. Our rationale for this is Influence Tactics, Respect and Human Resource Decisions 5 two-fold. First, flexible work arrangements are an increasingly important HR practice associated with employee well-being, engagement and commitment (Grzywacz, Carlson, & Shulkin, 2008;Richman et al., 2008). Increasingly, research highlights the need to distinguish between formal and informal arrangements that enable employees to access FWAs (De Menezes & Kelliher 2017). Furthermore, that most FWAs are decided by the supervisor (Troup & Rose, 2012). Supervisors have been found to act as "gatekeepers" to FWAs and can decide to restrict access even when formal FWA policies exist in an organization (Kossek et al., 2010). In the absence of formal FWA policies, supervisors also use their discretion in determining whether to allow employees to access FWA (Rousseau, Ho, & Greenberg, 2006). Consequently, there are calls to gain a better understanding of factors that might affect a supervisor's decision-making in this area (Poelmans & Beham, 2008). Investigating upward influence tactic effects can provide new insights into this decision-making process. Second, studies have suggested that the specific work outcome of interest (i.e. promotion, job performance ratings etc.) is likely to affect the relationship between influence tactics and work outcomes (Lee et al., 2017). For example, extrinsic success factors such as promotion and salary increases, are subject to a range of extraneous conditions outside the supervisor's control. Such conditions may therefore significantly limit the effects of relational quality (mutual respect) as an explanatory mechanism for upward influence tactic effects. However, both job performance ratings and granting FWAs fall more directly under the supervisor's control.
These outcomes are therefore more likely explained by relational quality mechanisms.
Testing the effects of mutual respect as a mediator between influence tactics on both these key work outcomes thus offers an opportunity to gain stronger empirical support for our posited theoretical model. Our findings contribute to the HR literature in furthering our understanding of individual (such as influence tactics) and contextual (such as mutual

Arrangements
The taxonomy developed by Shreisheim & Hinkin (1990) consisting of six upward influence tactics, has been the mostly widely adopted in research studies to date.
These authors categorised upward influence tactics as (1) Rational persuasion, where employees use logical arguments to convince their supervisors to follow a suggested course of action; (2) Ingratiation, where employees flatter their supervisors with compliments in order to receive more favorable treatment or to be thought of in a more positive manner; (3) Exchange of benefits, where the employee seeks to obtain a favorable outcome by agreeing to trade something with their supervisor in return; (4) Upward appeal, where the employee attempts to influence the behavior or decision of their supervisor by appealing to a higher authority; (5) Coalition, where the employee seeks the support of others as a means to apply pressure that the supervisor agree or acquiesce to the employee´s wishes; (6) Assertiveness, where the employee adopts a confrontational or persistent stance to influence the supervisor to agree a particular request. More recently, researchers have added a further influence tactic to this taxonomy called (7) self-promotion. This tactic is used by the employee to appear competent in the eyes of the supervisor. It differs from ingratiation which captures instead an employee´s attempt to manage their supervisor´s impression of them as likeable (Ferris et al., 1994).
To date, research has not explored whether these influence tactics might similarly affect supervisor decisions to grant flexible working arrangements (FWAs) to their employees. Flexible work arrangements are "the ability of workers to make choices influencing when, where, and for how long they engage in work-related tasks" (Hill et al., 2008, p152). Research has shown FWAs to confer significant benefits to both organizations and employees. This has included increased job satisfaction, job performance, reduced absenteeism, higher retention rates as well as reduced work-family conflict (Combs et al., 2006;Shockley & Allen, 2007;Stavrou & Kilaniotis, 2010). As in other areas of HR policy implementation (Sikora & Ferris, 2014), line managers play a key role in how flexible working arrangements are implemented, as well as their utilisation by employees (De Sivatte & Guadamillas, 2011;McCarthy, Darcy & Grady, 2010). Not least since supervisors are often responsible for the approval of flexible work requests, particularly in relation to schedule changes and flexibility (Lauzun et al., 2010, Poelmans & Beham, 2008. Indeed, some argue that the decision-making of supervisors Influence Tactics, Respect and Human Resource Decisions 8 in this area is far more important than other more formal means of organizational support for flexible working (Behson, 2005).
The use of upward influence tactics may be particularly significant, given that an employee's skills in bargaining and negotiation have been suggested as influencing whether supervisors grant them flexible working arrangements, or what are increasingly referred to as idiosyncratic deals (Rousseau, Ho & Greenberg, 2006). A survey of employees in the German civil service for example, found that employees' use of personal initiative was positively associated with negotiating idiosyncratic deals (Hornung, Rousseau & Glaser, 2008). Based upon a social information processing perspective (Srull & Wyer, 1989), we might reasonably expect to find relationships between employees' use of influence tactics and supervisor decisions regarding flexible work arrangements, similar to those previously found with respect to job performance ratings. We therefore hypothesize: Hypothesis 1a: The upward influence tactics of Rationality, Ingratiation and Selfpromotion will be positively associated with Job Performance Ratings.
Hypothesis 1b: The upward influence tactics of Rationality, Ingratiation and Self-Promotion will be positively associated with Flexible Work Arrangements.

Respect
Respect has been identified as one of the most important aspects of an individual's work relationship, even surpassing job security and salary (Cleveland, Byrne & Cavanagh, 2015;Rogers & Ashforth, 2017;Van Quaquebeke, Zenker, & Eckloff, 2008). Recently, there have been major developments in our understanding of the concept of respect that has highlighted differing forms of the construct, referred to as appraisal respect and recognition respect (Darwall, 1977). Honneth's (Fraser & Honneth, 2003;Honneth, 1995) notion of recognition respect distinguishes this form of respect as that which conveys acceptance and recognition as an equal, drawing upon ideas from Kant (1964) on positive self-regard. Recognition respect corresponds with a belief that all individuals should be treated equally and fairly with dignity. This form of respect denotes individual needs for due regard, based on their humanity. This contrasts with the appraisal form of respect, which instead is dependent upon an appreciation of merit, worth or status. This corresponds with the idea that we earn respect, perhaps due to particular characteristics or abilities of an individual being valued or highly regarded.
Both forms of respect have been shown to bring about feelings of self-worth and selfesteem, but are recognised as each being distinctive (Simon & Grabow, 2014;Simon, Grabow, & Bohme, 2015).
How group members behave towards an individual in terms of fairness, and status are key signals that an individual feels respected by them (Anderson, et al., 2001;Tyler & Blader, 2000). Consequently, individuals will then identify more strongly with the group and undertake group serving behaviors (Smith, Tyler & Huo, 2003;Tyler & Blader, 2003). There is an increasing acceptance that although both appraisal and recognition forms of respect can co-exist, they need not necessarily be congruent. Grover (2013) highlights that an individual may be bullied (an absence of recognition respect), despite a bully having an appreciation of (or indeed because of) their skills and qualities. How these different forms of respect are co-produced between supervisors and their subordinates through how they treat and behave towards one another is therefore of chief interest. Clarke (2011), suggested a number of supervisor behaviors likely to predict different types of respect. Behaviors conveying concern and attentiveness, competence in task achievement, and recognition of employee achievement were posited to convey supervisor's appraisal respect of their employee. Behaviors that convey an appreciation for particular qualities or status, are also more likely to be associated with appraisal respect. Whilst displaying moral and ethical behaviors, fairness and integrity were posited to demonstrate recognition respect. Similarly, behaviors that convey polite and positive treatment (as opposed to being construed as rude) are thought to be more associated with recognition respect. Furthermore, behaviors that violate expectations of trust are thought to negatively affect recognition respect (Grover, 2013). Much of the theorizing though has tended to focus on the behaviors of supervisors in bringing about respectful relationships (Van Quaquebeke & Eckloff, 2009). Far less is known about how employee behaviors influence these two forms of respect.
Influence tactics comprise differing employee behaviors, some of which are likely to affect both forms of respect, although we might expect some differences too.
Tactics that convey a lack of respect for a supervisor's status or authority are likely to have a negative impact on appraisal respect. Similarly, tactics that suggest a lack of integrity or fairness by the employee should negatively impact recognition respect. Both coalition and upward appeal are tactics that undermine the authority or status of the supervisor, since they attempt to circumvent or undermine the supervisor's hierarchical position. Both these tactics challenge the supervisor's expectations of their decisionmaking powers, and likely result in a supervisor considering a violation of their psychological contract has occurred. This will give rise to feelings of unfairness, associated with perceptions of recognition respect. Previously, behaviors conveying a lack of respect (or disrespect) have been associated with perceptions of psychological contract breach (Richter, et al., 2016). Assertiveness (as defined earlier), refers to a confrontational tactic used by an employee to pressure the supervisor for a favorable outcome for the employee. Again, this is likely to be perceived as breaking expected norms of behavior through undermining the status/position of the supervisor.
Confrontation also implies breaking acceptable norms of behavior that convey a lack of care for the person, disregard and even abuse. This is also consistent with our understanding of the role of workplace norms in perceptions of incivility (Andersson & Pearson, 1999;Schilpzand, De Pater & Erez, 2016). We would expect this also to have a negative effect on recognition respect. Empirically, there has been limited research exploring the consequences of psychological contract breaches from the employer's perspective (Shore et al., 2004). The immediate supervisor is recognised as one agent representing the employer and their reactions to perceptions of employee psychological contract breach have started to attract interest (Takleab & Taylor, 2003). A recent study found that supervisors responded to such breaches through decreasing their socioemotional investments in the relationship they have with the employee. Chen, Tsui, and Zhong (2008) found this resulted in a significant decrease in relational quality (as captured through LMX). We might therefore expect similar negative effects on both mutual appraisal and mutual recognition respect. We therefore hypothesize:

Hypothesis 3: Coalition, Upward Appeal and Assertiveness will be negatively associated with both Mutual Recognition Respect and Mutual Appraisal Respect
Employees who use rational persuasion to influence supervisors base their arguments on logical arguments and reasoning. This is likely to be perceived by the supervisor as acting with integrity. Consequently, this tactic should be expected to have a positive effect on recognition respect. Employees who use rational arguments, drawing upon evidence to support their position, is also likely to be perceived as a positive quality or characteristic of the employee. This would be valued by the supervisor, and similarly likely to have a positive effect on appraisal respect.

Recognition Respect and Mutual Appraisal Respect
Employees who use ingratiation to influence their supervisors, engage in flattery and highlight qualities they perceive in the supervisor to be valued. They draw attention to either the supervisor´s competence, experience or personality, that they suggest sets them apart from others. This enhances their status. Based upon social identity theory, a considerable body of research shows that behaviors which communicate an individual's standing or status, corresponds to an individual feeling respected (appraisal respect) (Tyler & Blader, 2003). By contrast, the tactic of self-promotion involves the employee highlighting their own competence in various areas of job performance. This should result in a supervisor increasing the level of appraisal respect they feel towards their subordinate. Indeed, previous research has found that showing appreciation for an employee's competence is associated with employees feeling they are respected by their supervisors (Van Quaquebeke, & Eckloff, 2009). Both the supervisor feeling respected and respecting their subordinate, should together contribute to mutual appraisal respect.
A recent experimental study found that behaviors conveying warmth had a stronger effect than those conveying competence on perceptions of liking, whilst the opposite was the case for appraisal respect (Oleszkiewicz & Lachowicz-Tabaczek, 2016). Based Influence Tactics, Respect and Human Resource Decisions 13 on these status and competence arguments, we would expect both these influence tactics to positively affect mutual appraisal respect rather than mutual recognition respect.

Hypothesis 5: Ingratiation and Self-Promotion will be positively associated with Mutual
Appraisal Respect.
By contrast, employees who seek to influence their supervisors through trading sources of benefit, could be seen as failing to be performing their duties or undertaking their responsibilities as expected in fulfilment of their employment contract. In these instances, the supervisor is likely to perceive them as lacking integrity. In this instance, we would expect this tactic to have a negative impact on recognition respect. It is also possible that subordinates who use exchange tactics that involve explicit or implicit offers to provide a favor or benefit to the supervisor in return for helping them perform a task, may be perceived as less competent by their supervisor (Wayne & Ferris, 1999).
This being the case we would expect the following:

Hypothesis 6: Exchange of Benefits will be negatively associated with Mutual
Recognition Respect and Mutual Appraisal Respect.

Performance Ratings and Flexible Work Arrangements
That individuals strive to be respected in their relationships is well-documented and signals that individuals perceive themselves as socially included and valued (Bartel, Wrzesniewski, & Wiesenfeld, 2012;Bergsieker, Shelton, & Richeson, 2010). This is no less the case in workplace relationships, where being respected is highlighted as among the top values individuals want from colleagues (Van Quaquebeke et al., 2009). Social identity theory posits that individuals strive to maintain a positive self-image and that one's membership of, and position in a group contribute towards this self-image (Tajfel Turner, 1986). Feeling valued and included (as indicated by appraisal and recognition respect) means an individual will be motivated to undertake tasks and exert more effort on behalf of the group (Huo, Binning, & Molina, 2010). These dual aspects of both inclusion and value, have been found to be independent aspects of the social identification construct (Ellemers, et al., 2011). A key aspect of social identity theory, is that an individual will engage in behaviors as a means to maintain their self-esteem (Messick & Mackie, 1989). This draws upon ideas from the "pleasure principle" that posits individuals will seek to protect a positive self-view and strengthen their selfconcept (Wells, 2001), even when they already possess positive self-views (Brockner, 1988). This is consistent with findings that show individuals with low self-esteem are motivated by self-protection, whilst those with high self-esteem by self-enhancement (Baumeister, Tice & Hutton, 1989). Abrams & Hogg (1988) further argue that individuals will be motivated to enhance their self-esteem through engaging in in-group favouritism in order to protect and enhance their positive self-concept. A number of empirical studies have since shown this to be the case (Crocker, Blaine, & Luhtanen, 1993;De Cremer, 2001). Further support suggesting this link to favouritism can be found from recent research by Ng (2016). Drawing upon social exchange theory, he found that employees responded to perceptions of workplace respect by becoming more positively embedded in the organization and that this was mediated by employee gratitude to the organization. We posit that in applying a similar line of thinking at the level of supervisor-employee relationship, then supervisors will respond to the respect they receive by showing gratitude towards their subordinate through favoritism. This should extend to the supervisor providing more favorable evaluations of employees that affect their human resource decision-making. Specifically, more favorable job performance ratings and flexible work arrangements. In relation to the latter, research has also found employees report that supervisors who grant them FWAs feel they are more respected by them (Koiviso & Rice, 2016). Subordinates who receive respect from their supervisors should respond with respect towards their supervisor as part of mutual social exchange (Clarke, 2011). Granting FWAs to subordinates would therefore seem likely associated with mutual levels of respect. Elsewhere, a relationship between a supervisor's positive self-concept and the granting of FWAs to subordinates has also been suggested as a possible explanation to account for variations in access to FWAs.
Sweet, Pitt-Catspouphes & Boone-James (2017) drew upon Conservation of Resources theory (Hobfoll, 2001) to explain managers' decisions as to whether to grant FWAs to a specific employee. This theory posits that how individuals respond to job demands is associated with the amount of job resources they have available, and that they will seek to preserve the resources they value. They argued that granting FWAs to some employees might be perceived by supervisors as them having to incur personal costs.
These costs arise through activities such as covering tasks, rescheduling work processes, and even potentially damaging longer term career prospects if their work unit fails to meet targets. They suggest that supervisors with a highly positive self-concept (such as that arising due to the respect they receive) are more likely to use proactive coping strategies when faced with task demands such as these. Consequently, they are more likely to consider they possess the resources to deal with these stressors (Brown, Jones, & Leigh, 2005). The finding that individuals who are respected are also perceived as being trustworthy and competent (Smith & Tyler, 1997) is likely also to strengthen these self-beliefs. It follows then, that mutual respect between supervisors and subordinates should be positively associated with supervisors granting FWAs as well as more positive job performance ratings.

The Significance of Mutuality
Similar to recent developments in the trust literature (Dass & Kumar, 2011;De Jong & Dirks, 2012), respect has been posited as a relational phenomenon rather than a property associated with an individual. In this sense, respect should not be thought of as a psychological state measured at the individual level, but instead is more properly measured at the dyadic level of analysis. It is the differences between groups (i.e. between supervisor-employee dyads) rather than within groups (between supervisors and employees) that is of interest here in examining relationships between respect and our posited outcomes. Similar to Clarke & Mahadi, (2017), we also suggest that mutuality in respect gives rise to additional benefits. A number of studies have found mutuality in relationships to bring about a number of positive psychological outcomes associated with self-worth and self-validation, as well as enhance coping mechanisms (Coyne & Bolger, 1990;Genero et al., 1992). One explanation for this is that mutuality implies a high degree of acceptance by the other party. This fosters a positive social identity, which then brings about enhanced feelings of self-esteem (Smith et al., 1988). This is consistent with research that has shown acceptance in social groups to be associated with self-esteem (Baumeister & Leary, 1995). Interdependence theory for example, suggests that individuals strive for closeness in relationships so as to maximise the exchange of pleasurable resources (rewards), whilst minimise any losses or costs including social and emotional resources (Kelley & Thibaut, 1978;Rusbult & Van Lange, 2003). The bounded reciprocity hypothesis (Gaertner & Insko, 2000), also suggests that perceptions of reciprocity are influenced not only by social categorization, but also by perceptions of interdependence. Based on early ideas posited by Rabbie, Schot and Visser (1989), this suggests that the level of resources allocated by someone to others, is influenced by the extent to which they perceive themselves dependent upon them for achieving particular outcomes. Research has found empirical support for both processes operating in (Stroebe, Lodewijkx, Spears 2005). Mutuality should therefore confer additional benefits in contributing to a supervisor's self-esteem, thus further enhancing in-group favoritism. We therefore hypothesize: The above hypotheses (3-7b) are represented diagrammatically in figure 2.

Procedure
Using personal contacts with HR Directors, we contacted thirty-one organizations in Saudi Arabia from both the public and private sectors who agreed to participate in the study. Senior managers then provided the names of matched independent supervisor-subordinate dyads from across their departments who were coded to assist matching on-line survey responses. The survey items were initially constructed in English and then translated into Arabic using the back-translation method (Brislin, 1970). A few inconsistencies appeared which were then resolved. This centred on some words in the initial translation and back translation that were different, but had the same meaning in Arabic. Once similar meaning was established by the research team, we proceeded with data collection. We provided identity codes to 680 matched pairs (1360 individuals) and received responses from 446 supervisors and 478 employees, a total response rate of all subjects of 68% (924). After matching responses, our final sample consisted of 389 supervisor-subordinate dyads, which is a dyad response rate of 52.2%. We ensured that all dyads were completely independent from one another (i.e. that each supervisor and subordinate appeared in only one dyad). were educated to undergraduate level, with a further 4.9% possessing a post-graduate qualification. Approximately 46% had been working for their organization for 7 years or more. Approximately 64% had been supervised by their current line manager for 2 years or less, whilst 25% between 3 to 6 years. Finally, the majority of our leader-subordinate dyads were working in public sector organizations (63%), with the private sector (35%) and non-profit sector (2%) accounting for the remainder of our sample. Public sector dyads were drawn from the following workplaces: healthcare 117 (30.1%), education 60 (30.1%) and environment & planning 71 (18.4%). Those from the private sector were financial and insurance 43 (11%), retail 35 (9.0%), healthcare 26 (6.7%), telecommunications 24 (6.2%) and business consultancy 7 (1.8%). Finally, 6 dyads (2%) of our sample were drawn from two charitable foundations.

Measures
Upward Influence Tactics.
We obtained measures of employees´ use of upward influence tactics from their line manager, using the 18-item scale by Hinkin and Schriesheim (1990) to measure six influence tactics. We added four items from the Ansari et al., (2007) Self-Promotion scale.
Line managers were asked to rate how often their employees used each tactic with them in the past 6 months using a 5-point Likert scale (1=never uses this tactic, 5=usually uses this tactic.

Job Performance Ratings.
We used 7-item in-role performance scale (Williams & Anderson 1991) and asked managers to rate their employee's performance on a 5-point scale.

Flexible Working Arrangements
Subordinates indicated the extent of their flexible working on a 5-point scale, using the 4-item flexible working scale by Hill et al., (2001).

Mutual Recognition Respect
We used the 8-item mutual recognition respect measure developed by Clarke & Mahadi (2017) with a 7-point scale, and collected measures from both managers and their subordinates.

Mutual Appraisal Respect
We used the 3-item professional respect developed by Liden & Maslyn (1998) to measure mutual appraisal respect changing the wording slightly depending on the target. Measures were collected from both managers and subordinates using a 7-point scale.

Control Variables
The similarity-attraction paradigm (Byrne 1971) suggests individuals with similar demographics tend to like each other more and this has also empirically been found to be the case (Varma and Stroh, 2001 Teachout, 1995). Given that we are examining the mediating effects of relationship quality (mutual respect) we decided to include both gender similarity and length of relationship as control variables in our study.

Data Analysis
We used AMOS Ver18 and SPSS Ver21 statistical software and followed procedures suggested by Preacher, Rucker & Hayes (2007) for testing mediation, having obtained a macro from the authors (http://www.afhayes.com/spss-sas-and-mplus-macrosand-code.html). This allowed us to apply bootstrapping (5000) and obtain bias-corrected confidence intervals in testing our posited mediated relationships (Edwards & Lambert, 2007). This approach is suggested to provide results for direct and indirect effects similar to those expected using structural equation modelling (Hayes, 2013, p196-197).

Confirmatory factor analysis
To demonstrate the discriminant validity of our measures, we undertook a confirmatory factor analysis (AMOS 18) comparing the chi-squared and goodness of fit indices (Anderson & Gerbing, 1988) between four separate, estimated models. In our first model, we loaded both mutual recognition respect and mutual appraisal respect scale items on to one factor, items from both our outcome variables (job performance ratings and flexible work arrangements) on to a second factor, 3 influence tactics (ingratiation, rational persuasion and self-promotion) on to a third factor, and the items from our remaining influence tactics (assertiveness, exchange of benefits, coalition, and upward appeal) on to a fourth factor. The estimated model demonstrated a very poor fit to the data (χ2 = 4796.33, df = 939; p<.000; CFI =.65, GFI = .60, RMSEA = .10). In the second of our estimated models, we followed the same loading of items on to factors as previously, save loading each of the influence tactic scale items on to its corresponding factor. Again, the model showed a poor fit to the data (χ2 = 3197.25, df = 909; p<.000; CFI =.80, GFI = .71, RMSEA = .08). In our third estimated model we loaded all items on to its corresponding factor, except job performance ratings and flexible work arrangements which we loaded on to one factor. This model demonstrated a much better fit to the data than previous models. (χ2 = 2178.88, df = 890; p<.000; CFI =.82, GFI = .86, RMSEA = .07). Finally, we loaded each item onto its corresponding factor creating a 13 factor estimated model. This measurement model demonstrated the best fit to our data (χ2 = 3319.28, df = 1352; p<.000; CFI =.90, GFI = .89, RMSEA = .06). This supported the discriminant validity of our measures.

Aggregating Data to Dyad Level
We employed two approaches accepted in the literature to determine whether we were justified in aggregating our data to the dyad level to determine mutuality. First, we obtained intra-class correlations (ICC1) for both appraisal and recognition respect. The intra-class correlation indicates the variation that can be accounted for by group membership (Bliese 2000). Our ICC1 values were Appraisal Respect, (ICC1 = .77, F = 4.36); and Mutual Recognition Respect (ICC1 = .92, F = 1231). Both values were statistically significant (p< .001) and judged as indicating a high level of convergence in scores. We next estimated the degree of agreement between supervisors and their subordinates on items they rated for both appraisal and recognition respect by calculating the rwg statistic. This provides information on the degree of inter-rater agreement among raters using multiple item rating scales (James et al., 1984). The mean rwg(8) score we obtained for recognition respect was 0.97 and the mean rwg(3) score obtained for appraisal respect was 0.94. In addition we found 88% of our dyads for appraisal respect and 96% of our dyads for recognition respect had scores above the 0.70 criterion (Biemann, Cole, & Voepel, 2012 aggregation. We therefore calculated values for mutual recognition and mutual appraisal respect by using the mean of supervisor and subordinate scores on our respect measures.

[INSERT TABLE 1 ABOUT HERE]
Descriptive statistics Table 1 shows descriptive statistics, and inter-correlations for our study variables. This shows positive and significant relationships between ingratiation, rationality and self-promotion with job performance ratings. Whilst the four tactics, exchange of benefits, assertiveness, upward appeal and coalition, were negatively associated with job performance ratings. There are no significant relationships between any of the influence tactics and flexible working arrangements save assertiveness. Both mutual recognition respect and mutual appraisal respect were positively associated with both job performance ratings and flexible working arrangements. Three influence tactics (exchange of benefits, upward influence and coalition) are negatively associated with both mutual recognition respect and mutual appraisal respect. Whilst assertiveness is also negatively associated with mutual recognition respect. The upward influence tactic of rationality was positively associated with both forms of respect. Whilst selfpromotion also was positively associated with mutual appraisal respect. Finally, we observe a significant correlation (0.68) between mutual recognition respect and mutual appraisal respect. Previously, Simon, Grabow and Bohme (2015) supported the construct validity of these two forms of respect, through showing that equality-based respect (recognition respect) and achievement recognition (appraisal respect) each explained unique variance in an overall perception of what it means to be respected across the contexts of work, family and friends. Although distinct constructs, appraisal and recognition respect are part of the broader nomological network related to respect and we would therefore expect them to be significantly related. Other studies have similarly reported significant correlations between these two types of respect of .48 (Simon & Grabow 2012), .58 (Renger & Reese, 2017) and .75 (Clarke & Mahadi, 2017). Table 2 presents our results. This shows that rational persuasion (β= .23, p<.01), ingratiation (β= .06, p<.05) and self-promotion (β= .09, p<.01) were all positively associated with job performance ratings supporting hypothesis 1a. However relationships between rational persuasion (β= -.12, p<ns), ingratiation (β= -.02, p<ns) and self-promotion (β= .03, p<ns) and flexible work arrangements were all insignificant. Hypothesis 1b was therefore not supported. Significant negative relationships were found between exchange (β= -.14, p<.01), coalition (β= -.13, p<.01), upward appeal (β= -.17, p<.01) and assertiveness (β= -.10, p<.01) and job performance ratings supporting hypothesis 2a. However a significant relationship was found only between assertiveness and flexible work arrangements (β= -.11, p<.05), whilst relationships with exchange (β= .06, p<ns), coalition (β= .08, p<ns) and upward appeal (β= .04, p<ns) were all insignificant. Hypothesis 2b was therefore only partially supported. Coalition (β= -.04, p<.01), (Z= -2.49, p<.01) and upward appeal (β= -.06, p<.01), (Z= -4.03, p< .01) predicted mutual appraisal respect, and mutual appraisal respect predicted job performance ratings in each case respectively (β= .29, p<.01), (β= .27, p<.01).

Influence Tactics Direct and Indirect Effects [INSERT TABLE 2 ABOUT HERE]
However, the relationship between assertiveness and mutual appraisal respect was not significant (β= -.03, p<ns). Hypothesis 3 was therefore partially supported.

Discussion
Identifying factors influencing how supervisors make HR-related decisions and the processes involved is an area of key concern in the HR literature (Bernadin et al., 2016;Pichler, 2012). Whilst much of this work has focused on job performance ratings, we have extended this work further by also investigating supervisors' decisions regarding flexible working arrangements. Importantly, our findings contribute theoretically to our understanding of how social context affects these HR decisions and have important implications for HR practice.

Theoretical Implications
Although there now exists a major body of empirical literature showing significant relationships between employees´ use of upward influence tactics and supervisors´ HR decisions, far less work has sought to investigate the underlying theoretical mechanisms involved. Our findings offer support for mutual appraisal respect and mutual recognition respect between supervisors and their subordinates, as a potential mechanism whereby upward influence tactics can bring about their effects. This corresponds with growing evidence that relational quality seems to affect job performance ratings. A study by Alexander and Wilkins (1982) for example, showed relationship quality was a stronger predictor of job performance ratings than objective performance measures. Duarte, Goodson, & Klich (1994), found similar results using leader member exchange as a relational measure. This suggests supervisors may act more favorably towards employees with whom they share a better quality relationship, possibly motivated by some form of social reciprocity. Typically, professional respect (a form of appraisal respect) is included in measures of LMX alongside affect (liking), competence and contribution (Liden & Maslyn, 1998). It would seem that both liking and respect represent alternative mechanisms through which upward influence tactics bring about their effects. This being the case, LMX and mutual forms of respect (not captured in the LMX construct) may potentially have interactive effects. Our findings that mutual appraisal and mutual recognition respect influence supervisor ratings suggest that different aspects of relational quality may indeed contribute independent effects.
Our finding that both mutual appraisal and mutual recognition respect were positively associated with supervisor job performance ratings, is consistent with previous research suggesting that relational quality affects performance ratings (Levy & Williams, 2004). We also found both forms of respect to predict flexible working arrangements. Hass & Deseran (1981) have also suggested that supervisors granting FWAs to their employees symbolises a relationship characterised by strong socio-emotional bonds.
Previous research has highlighted the important role that supervisors play in granting access to FWA (Armstrong-Stassen & Schlosser 2010;Bardoel, 2003;Kossek, Lewis, & Hammer, 2010). This has found supervisor attitudes to flexible working, previous experience of managing employees with flexible work schedules, support for organizational FWA policies, and the extent of disruption to or effort involved for the supervisor as key factors influencing supervisor decisions (Julien, Sommerville, & Culp, 2011;Powell & Mainiero, 1999;Putnam, Myers, & Gailliard, 2014;Sweet, Pitt- Catsouphes & Boone James, 2017). Given that mutual appraisal respect is in part based on judgements of worthiness derived from abilities or competence, its significance here may reflect greater confidence by the supervisor that granting FWA will not result in them having to expend greater effort themselves. Previous research has also shown that supervisors rely far more on appraisals of competence in determining the trustworthiness of their subordinates (Knoll & Gill, 2011;Lapidot, Kark, & Shamir, 2007). This would correspond with studies suggesting that supervisors´ attributions of employee motives for requesting FWA affect their decisions on whether to grant it (Leslie, Park, & Mehng, 2012).
Our findings would also suggest that fears highlighted by Rousseau (2005), that employees receiving FWAs might result in them receiving lower job performance evaluations (as a result of being less visible for example) may not be justified. Our findings that both flexible work arrangements and job performance evaluations are similarly predicted by mutual appraisal respect might explain why this may not the case.
However, our finding that mutual recognition respect similarly predicted flexible working arrangements suggests that competence or trustworthiness attributions potentially arising from mutual appraisal respect cannot be the only explanation. Instead it offers some preliminary support at least, that enhancing supervisor self-esteem through respect, might represent an alternative mechanism to account why subordinates might receive more favorable supervisor HR decisions compared to others.
Here we drew upon social identity theory (Lind 2001) to account for the effects of mutual respect. We posited that the social categorization effects brought about by both mutual appraisal and mutual recognition respect within a supervisor-subordinate relationship, will enhance a supervisor´s self-esteem, which leads to favoritism as a form of social reciprocity. Both of these forms of respect are found to foster identification with one's immediate work group (Renger & Simon, 2011;Simon & Sturmer 2003). Theories of self-construal suggest that an individual´s sense of the self or self-concept, is derived through the interactions they experience with significant others (Shotter & Gergen, 1989).
High levels of mutual recognition and mutual appraisal respect convey acceptance and fair treatment and contribute to a positive social identity. This then gives rise to positive affect and judgements of self-worth and self-esteem (Baumeister & Tice, 1990, Smith et al., 1988. Previously, studies have found these forms of respect to foster group identification and group serving behaviors (Ellemers et al., 2011;Tyler & Blader, 2003).
We suggested that in supervisor-subordinate relationships, these forms of respect will motivate supervisors to act more favorably towards subordinates (Abrams & Hogg 1988;Crocker et al., 1993;De Cremer, 2001). This also explains why supervisors may be disposed to act more favorably towards some employees when rating job performance and responding to FWA requests. Our findings would seem to offer some preliminary support for these arguments.
We found similarities and some differences in the extent to which upward influence tactics predicted these two forms of respect in the relationships we investigated.
We found 3 tactics (coalition, upward appeal and exchange of benefits) were all negatively associated with both mutual recognition respect and mutual appraisal respect.
Assertiveness by contrast, was only negatively associated with mutual recognition respect.
Our finding that assertiveness was only significantly associated with mutual recognition respect and not mutual appraisal respect was surprising, since confrontational approaches associated with this tactic similarly, can be thought of as undermining the authority (or status) of the supervisor as well as showing disrespect for the person. It is interesting that of the tactics typically found to result in negative responses from supervisors, this is the only tactic that does not involve the employee breaking the chain of command or enlisting the support of others to influence a supervisor´s decision making. It may be that these former tactics might have a more significant effect on supervisor´s perceptions of psychological contract violations, thus effecting status perceptions and consequently mutual appraisal respect (Lester et al., 2003). Whereas confrontational behaviors captured in assertiveness tactics are more keenly perceived as violations of the self, thus affecting mutual recognition respect.
We found rational persuasion to be positively associated with both mutual recognition respect and mutual appraisal respect, whilst self-promotion was only positively associated with mutual appraisal respect. Self-promotion involves the employee highlighting their achievements and abilities to their supervisor. Fragale & Grant (2015) suggested that the effects of self-promotion might be explained through a functionalist perspective of status conferral. They argue that individuals in groups and organizations best able to assist in achieving collective goals, are those often granted higher status. Individuals who self-promote point out how competent they are, and are therefore perceived as important to the supervisor for achieving tasks. Consequently, status is conferred through high value signals such as positive job evaluations as well as financial rewards (Abele & Wojciszke, 2007;Fragale, 2006). This would seem to correspond with our finding here that competence-oriented behaviors (conveyed through influence tactics of self-promotion) appear to have a significant effect on mutual appraisal but not on mutual recognition respect.
Our findings that assertiveness only predicted mutual recognition respect whilst self-promotion only predicted mutual appraisal respect also offers further evidence suggesting the distinctive nature of these two forms of respect. Although studies Influence Tactics, Respect and Human Resource Decisions 30 examining the recognition form of respect are relatively recent compared to appraisal respect, the distinctiveness of these two constructs in differentially predicting outcomes is growing. Clarke & Mahadi (2017), found that whilst mutual appraisal respect and mutual recognition respect both predicted employee job performance, only recognition respect predicted employee well-being. More recently, Renger & Reese (2017) showed that whereas the recognition form of respect (equality-based respect), predicted global identity, appraisal respect did not. In this instance we have shown differences between two influence tactics in their predictive relationships with these two forms of respect.
Within the HRM literature, the concept of respect has been highlighted as central to effective employee relations and healthy interpersonal relationships (Ng, 2016). Much of this work has tended to utilise global measures of respect that fails to operationalise differences between appraisal and recognition respect. Our findings would therefore support incorporating both these alternative forms of respect in research to gain a better understanding of their value in predicting particular outcomes and how their antecedents might differ.
Finally, our findings showed a number of direct effects for influence tactics on job performance ratings and flexible work arrangements, as well as indirect effects through our mutual respect measures. Consistent with previous research, we found the tactics of rational persuasion, ingratiation and self-promotion to be positively associated with job performance ratings. Whilst exchange of benefits, coalition, upward appeal and assertiveness were found to have negative effects (Rao et al., 1995;Yukl & Tracey, 1992).
Although there have been a number of country comparison studies (Ralston et al., 2002(Ralston et al., , 2005Yukl et al., 2003;Botero et al., 2012), this has overwhelmingly investigated and found differences between cultures in terms of their preference for the use of particular upward influence tactics. Ralston et al., (1993) for example, found that Hong Kong managers were more likely to employ tactics that involved controlling information within their network compared to Americans. By contrast, American managers reported the use of image management tactics far more frequently. Yukl et al., (2003) conducted two studies, the first study comparing American, Swiss and Chinese managers. The second American, Hong Kong and Chinese managers. The results showed that Western managers believed that rational persuasion, inspirational appeal and coalition tactics were more effective than Chinese managers. By contrast, Chinese managers believed that appeals to authoritative bodies might be a more effective tactic. However, despite differences found in the preference for particular upward influence tactics, few studies have examined whether the effects of upward influence tactics on job performance ratings are consistent across cultures. Most studies examining relationships here have been overwhelmingly conducted in the United States (Higgins et al., 2003). However, Su (2010) conducted a study in Taiwan and found assertiveness and rational persuasion had similar effects on job performance ratings to those found among US managers. Pandey & Singh (1987) studied the effects of an upward influence tactic called other enhancement (similar to the notion of ingratiation), and again found this was positively associated with job performance evaluations. This is the first study to our knowledge that has examined the direct effects of upward influence tactics on job performance ratings in a Saudi Arabian context, and our findings appear consistent with those reported in previous studies in the US.
In relation to FWA however, we found significant, direct effects only for assertiveness. This suggests that how supervisors reach decisions in differing areas of HR responsibility such as job performance ratings and granting flexible work arrangements, may be far more complex than any one theoretical explanation can account for. Further, that they are likely to be subject to decision-domain specific individual and situational contingencies. Nevertheless, the significant effects we found for employee upward influence tactics affecting these decisions suggests employees should not be treated as merely passive participants in how HR policies are implemented in organizations.
Although, much research on HRM and its organizational effects is dominated by macro and systems level perspectives (Boxall & Macky, 2007;Paauwe & Boselie, 2005), investigating how policies are implemented at an individual level has considerable value.
Employees are increasingly engaging in behaviors to shape their job roles, how they perform their duties as well as their work conditions (Rousseau, 2001, Wrzesniewski & Dutton, 2001. This arises through their ability to shape "idiosyncratic deals" with their organizations most commonly with their immediate line managers. One of the most common forms of these deals focuses on flexible working arrangements (Rousseau, 2005).
Employees also take a far more active role in shaping their supervisor´s impressions of themselves to receive other more favorable outcomes such as job performance ratings.
More broadly however, these effects may well impact on other employee perceptions of fairness of the HRM system in an organization more generally, i.e, the distribution rules by which individuals understand effort-reward outcomes. Bowen and Ostroff, (2004) highlight employee perceptions of fairness as an important indicator of "HRM system strength" and that this is likely to play a role in understanding the HRM-Performance relationship. This suggests we need to minimize bias that can occur when supervisors make HR decisions in these areas.

Practical Implications
Increasingly, social context is recognised as affecting the performance appraisal process (Erdogan, 2002;Ferris et al., 2008;Levy & Williams 2004). Our findings suggest that mutual appraisal and mutual recognition respect between supervisors and their subordinates, may contribute to bias in supervisor ratings. Given that supervisor job performance ratings are often used to inform an employee's promotion and reward, this clearly adds a further layer of complexity for organizations seeking to minimize rater bias.
Much effort has been directed to improving the psychometric properties of rating scales and providing rater training (Roch et al., 2012), given that supervisor competence in the area contributes to bias (Bernadin, et al., 2016;Hoffman et al., 2010). However, these interventions are likely to be less effective in addressing respect between supervisors and their subordinates as a potential source of bias. Instead, organizations should rely on multiple raters of subordinates' job performance. This is likely to have a greater impact in addressing bias originating from relationship differences (Murphy, Cleveland, & Hanscom, 2018;Tetlock & Gardner 2015). A chief implication for managers from this study is that they should be more aware of how differing relationships with their subordinates might affect their HR decision-making. They therefore need to consciously take steps to minimize bias (Russo & Shoemaker 2001).
Our study also suggests some practical implications for employees. Depending upon the particular influence tactic used, this can have either positive or negative effects.
The tactics of rational persuasion, and self-promotion were found to have positive effects, and so employees should be encouraged to incorporate these tactics when attempting to influence or negotiate with their supervisors. Employees should also be aware that the use of "harder" forms of influence tactic may well result in them receiving less favorable outcomes in these areas. Organizations often provide training in negotiation skills for their employees which include the use of influence tactics. Incorporating awareness of the effects of upward influence tactics and how to implement those tactics found to have more positive effects, would therefore be worthwhile.

Limitations and Future Research
The results of our study should be considered in the context of its limitations. First, is that our cross-sectional research design precludes emphatic causal inferences. Whilst the order of our variables make sense theoretically, we cannot preclude a reverse causal direction. Studies have previously found relationships between leader-member exchange (an indicator of relationship quality) and upward influence tactics on the basis that social context affects employee choices on which influence tactics to use with their supervisors (Deluga & Perry, 1991;Olufowote, Miller & Wilson, 2005). Recently, supervisors granting FWAs to employees has been found to enhance their feelings of respect (Kovisto & Rice, 2016). It is possible then, that reciprocal influences exist. Future research that employs a time-lagged design are therefore necessary. Next, we used supervisor ratings of employee influence tactics. Again, this was consistent with our view and findings from previous research, that it is supervisor perceptions of employee behaviors that are likely to affect HR decisions such as job performance ratings (Allen & Rush 1998;Rotundo & Sackett 2002). It should also be noted that the measure of upward influence tactics we used ) was specifically developed for use as a self-report measure. Both the source of reporting as well as the measure used, have been found to moderate relationships between influence tactics and outcomes. Using employee selfratings of upward influence tactics is therefore likely to have produced different results to those that we obtained. Other measures of influence tactics have more recently also sought to incorporate other forms of influence, such as collaboration and consultation.
These reflect changing values and expectations more broadly of today's workforce. These are not captured in the  typology.
A number of contextual factors may also limit the generalizability of our findings.
We conducted our study in Saudi Arabia. Culture has been found to have a significant impact on both the choice and effects of upward influence tactics (Botero, Foste, & Pace, Ralston et al., 2005). In Saudi Arabia, respect for authority and status reflects its high power distance (Tlaiss & Elamin, 2015). This may well explain the results we found for assertiveness and self-promotion, in differentially predicting mutual recognition respect and mutual appraisal respect. Culture has also been found to influence the introduction as well as uptake of flexible working arrangements by employees (Kassinis & Stavrou, 2013;Masuda et al., 2012). It has been suggested that individuals in collectivist countries are less likely to use FWAs compared to those in individualist countries, as the latter place far greater importance on individual rights. Similarly, since FWAs suggest individuals are seeking greater control over their work conditions, they are also more likely to occur in countries characterised as having a low power distance (Peretz, Fried, & Levi 2018). Saudia Arabia is categorised as both collectivist and possessing a high power distance. This could explain why we found weaker relationships between our respect measures and flexible working arrangements than we did for job performance ratings. Future research should therefore aim to examine the mediating effects of both forms of respect with influence tactics and flexible working arrangements in other cultural contexts, particularly those characterised as having low power distance/individualism.
The characteristics of our sample may also have affected our findings. Over 60% of our dyads were drawn from the public sector. In a recent meta-analyses of influence tactics, Lee et al., (2017) found that the study setting did have an effect on the relationships found, although this was far more pronounced (stronger effects) for relationship-oriented outcomes (such as trust and LMX) then on task-related outcomes (such as job performance evaluations). In relation to the latter, the authors were only able to examine the effects of ingratiation and coalition, so our knowledge here remains Further work is also needed to understand more fully why these differing types of respect actually influence supervisor´s HR decisions. Our findings are consistent with previous studies, suggesting aspects of relational quality (including liking, perceptions of competence, and LMX) may predict job performance ratings. However, we posited a particular role for respect indices as this has been found to account for favoritism based on the benefits to self-esteem. However, we never included self-esteem as an intervening variable in our study. Future research that includes supervisor self-esteem as a mediator between mutual appraisal and mutual recognition respect and the outcome variables we studied, would offer stronger support for the respect-supervisor decision bias link we theorized here.
Future research might also focus on the extent to which there might be moderators of the upward influence tactic-mutual respect relationship. Demographic and other similarity factors between supervisors and their subordinates have been found to predict relationship quality. We might therefore expect stronger relationships between influence tactics and mutual appraisal and mutual recognition respect, the more they share personal similarities. Finally, future studies should investigate potential moderators of the relationships between both mutual appraisal respect and mutual recognition respect and job performance ratings. Given that mutuality implies a degree of dependence on one another shared by both parties in a relationship, then perceptions of interdependence may well be a potential moderator (Gaertner & Insko, 2000). Similarly, personality variables that might attenuate a supervisor's response to respect in the workplace such as selfregulation, negative affect or narcissism might also have moderating effects.

Conclusions
Our study has answered calls for more research on the underlying mechanisms by which influence tactics bring about their effects (Castro et al., 2003;Higgins et al., 2003;Yukl et al., 2005). We drew upon social identity theory, in positing two distinct forms of respect, mutual appraisal respect and mutual recognition respect as mediators of relationships between a number of upward influence tactics and supervisor HR decisions.
We reasoned that both forms of respect would result in a supervisor showing more favoritism towards their subordinates, and found some initial support for our reasoning.
Importantly, both mutual appraisal respect and mutual recognition respect were found to mediate relationships between a number of upward influence tactics and both job performance ratings and granting flexible working arrangements to subordinates. The influence tactics of coalition, upward appeal, exchange of benefits and rational persuasion predicted both forms of respect. However, assertiveness only predicted mutual recognition respect and self-promotion only predicted mutual appraisal respect. This supports the distinctive nature of these two forms of respect. Our findings suggest these forms of mutual respect capture an important aspect of relational quality beyond liking, which can account for more positive or favorable HR decisions by supervisors.