Recruitment agencies and unethical client requests: the ‘loyal matchmaker’ dilemma

Abstract This research focuses on (non-)compliance with unethical requests in the context of external recruitment specialists and their susceptibility to the ‘loyal matchmaker’ dilemma. A total of 176 Israeli recruitment specialists were presented with ethical dilemmas in which a client makes an unethical request. The focal point of ethical obligation was measured based on the Basic Human Values theory. Results indicated direct and indirect connections with the readiness to comply with unethical requests. Conformism with agency rules exhibited a positive connection with compliance. Benevolence and universalism led to decreased readiness to comply with unethical requests, although this behaviour was contingent on the particular situation. Recruitment specialists who are high in self-direction altered their readiness to comply based on the stakeholders. By highlighting the difficulties faced by recruitment specialists vis-à-vis rooting their obligations to clients and job-seeking candidates in ethical standards, the results of this study have important theoretical and practical implications.

to a specific appearance, which results in unethical hiring practices. The recruitment specialist is thus faced with an ethical dilemma: they can disobey the client's demand (and give the firm the résumé of a candidate who does not possess the desired ethnic traits but is qualified) or they can accede to the unethical request (i.e., screen out candidates for the desired ethnic profile). This problem is an example of the 'loyal matchmaker' dilemma in employment discrimination law (Seipel, 2017).
In the literature, less attention has been given to ethical dilemmas that confront external recruitment specialists. Sweeney (2011) found that recruitment agencies use racial stereotypes in describing, analysing and matching job applicants with employees, even though this practice is prohibited by law. As these agencies function as gatekeepers between candidate and organisation, their behaviours may deepen discrimination. However, little is known about the decision-making of recruitment specialists -their definition of ethical boundaries and what constitutes unethical requests (Linehan and O'Brien, 2017). In this research, we will explore the readiness of recruitment specialists to obey unethical demands in order to ensure a transactional arrangement between candidate and organisation.
We based our arguments on Treviño's (1986) ethical model, which proposed that ethical decision-making includes both individual and situational aspects. He noted ego strength, field dependence and locus of control as individual factors that impact ethical decision-making. Previous studies have found that HR managers are high in agreeableness, teamwork and customer service orientation, compared to other occupation types (Jamshidi et al., 2012;Lounsbury et al., 2008). Thus, we wanted to focus on the potential clash between the need to satisfy the organisational client and the individual candidate in a recruitment specialist framework, that is, the recruitment specialist's focal point of obligation to their agency. Are they more committed to their agency, the welfare of society or their own self-interest? We argue that the ethical ramifications of this focal point of obligation influence recruitment specialists´ readiness to conform to the loyal matchmaker dilemma, as they inform the desire to satisfy and cooperate with other's demands.
To evaluate this focal point obligation, we examined personal values. Personal values have hitherto been revealed to influence HR managers' ethical decision-making and behaviours (Fisher and Lovell, 2003). Here, we rely on Schwartz's basic human values theory (Schwartz, 1992(Schwartz, , 2012 and focus on four types of values that represent different focal points of obligation for the employee: conformity, benevolence, universalism and selfdirection. Valuing conformity translates into self-discipline and conformity with a firm's rules and regulations. On the other hand, the value of self-direction implies a focus on an individual's own motivation, independent thought and action. Both benevolence and universalism values are concentrated on the welfare of others (Schwartz, 2012).
With that, working in recruitment agencies does not imply that the recruitment specialists are driven by social motivation, even if they are high in benevolence and universalism values. While some of them adopt a welfare tradition towards employees, others see HR in terms of the development and maintenance of strategic business partners. These specialists serve as agents of firms' owners, while trying to satisfy both the candidates and the organisational clients. In our study, we suggest that recruitment specialists' motivation to comply with unethical requests from clients may be associated with their values, and the order of importance, and not necessarily from perceiving recruitment agencies as aiming to serve social benefits.
Only a few studies have concentrated on the problem of the loyal matchmaker dilemma in recruitment agencies (Seipel, 2017), as well as focal points of ethical obligations vis-à-vis altering responses to unethical client requests (e.g., Van Kenhove, Vermeir and Verniers, 2001). As such, the present study contributes to ethical theory, organisational behaviour and HR theories. Moreover, our conclusions can offer practical solutions for HR personnel facing ethical challenges.

THEORETICAL FRAMEWORK (Un)ethical work behaviours
Ethical business behaviour comprises organisational conformity to moral norms, rules, standards, principles and codes, as defined by the firm as guidelines for morally sound behaviours (Jones, 1991;Lewis, 1985). Generally, these codes are compatible with the moral standards of the society in which the organisation is located (Kaptein, 2008;Vardi and Weitz, 2004). While ethical behaviour indicates organisational adherence to moral standards, unethical behaviours represent violations of these social expectations (Jones, 1991).
Unethical work behaviours overlap with related deviant organisational behaviours such as counter-productivity, misbehaviour, antisocial actions and violation of regulations (Kaptein, 2011). Contrary to counterproductive behaviours, workplace deviance or antisocial behaviour, unethical work behaviours are not usually informed by deliberate intent to harm the organisation or its stakeholders. For example, while some managers consider promising unrealistic expectations from a product in an attempt to maximise sales, or improperly gathering confidential competitor information, as normative, others (e.g., the client or competitor) judge them as unethical deeds (Kaptein, 2008). In contrast, other forms of workplace deviance (e.g., office gossip, ignoring co-workers and avoiding volunteering for projects) may not necessarily fall under the rubric of unethical work behaviour -even if they do constitute forms of counterproductive work behaviours (Kish-Gephart et al., 2010).
Unethical behaviours are also distinct from illegal activities, although in some (but not all) cases, the former implicate the latter (Treviño and Nelson, 2016). The distinction between unethical and illegal resides in the formality of the action. While illegal and criminal behaviours violate official rules, unethical behaviours disrupt formal, informal and implicit organisational norms of behaviours (Jones, 1991;Kaptein, 2008;Kish-Gephart et al., 2010).

Factors that affect ethical decision-making among HR managers and recruitment specialists
Several theoretical models have mapped the mechanisms of ethical evaluation and behaviour (Jones, 1991;Kohlberg, 1981;Treviño, 1986). One familiar model is that put forward by Treviño (1986), based on Kohlberg's (1981) cognitive moral development (CMD) framework. Treviño's person-situation interactionist model posits that, following recognition of a situation as an ethical dilemma, judgment is moderated by individual and situational factors. These include ego strength and locus of control (e.g., individual factors) as well as job context and organisational culture. Indeed, LaMontagne (2016) found that, when confronting ethical dilemmas, HR managers rely on personal, situational and organisational factors to interpret the situation and to react. Specifically, HR managers highlight their personal values and professional identities as factors that contribute to their decisions. They describe their values as being fairness, justice and consideration of others' welfare. These values combine with their professional identity as caretakers and fixers -the people who 'makes things right' in the firm. Similarly, Fisher (1991: 246) found that, compared to accountants, HR specialists feel that they have to be involved in whistle-blowing situations and are the 'keepers of organizations' conscience'. These attributes indicate their obligation both to the firm and the employees, or to the welfare of society.
While internal HR managers have a clear professional and organisational identity, the current study explores borderline cases, such as recruitment specialists who work in external recruitment agencies, where responsibility extends both to the organisational clients and the candidates to whom they provide a service. Therefore, they are more vulnerable to the loyal matchmaker dilemma. This dilemma exists when a recruitment agency complies with, or acquiesces to, a discriminatory request by an organisation, because it sees itself as loyal to its organisational client. The effect of the loyal matchmaker is not merely theoretical. Several lawsuits have been filed against recruitment and staffing agencies, insisting on their liability in employee discrimination (Wears and Fisher, 2012). Therefore, this issue of the focal point of obligation has implications for the ethical decision-making process of recruitment specialists.
The current study suggests that the employee's particular focal point of obligation -whether that be the organisation, the welfare of society or their own self-interest -intervenes and affects recruitment specialists' decision-making tendencies, that is, it will impact how they react to ethical dilemmas. To measure the focal point of obligation, we rely on Schwartz's basic human values model (Schwartz, 1992).

Basic human values and the tendency to comply with ethical behaviours
The basic human values model (Schwartz, 1992) is one of the most recognised models for structuring the construct of human values. Schwartz's value theory (1992) suggests that universal human values are composed from 10 core attributes -common to all cultures, genders and ethnicities Shtudiner, 2016, 2021;Schwartz, 2016;Schwartz et al., 2012). Although universal, its hierarchy of importance differs between people. Therefore, one person ranks a certain value (e.g., benevolence) as the most important value for him, while another ranks the opposite value (e.g., hedonism) as the most important for her. The hierarchy order affects how people perceive situations and react to ethical dilemmas. Similarly, we assume that different value orderings will impact reactions towards organisational ethical dilemmas. To measure the focal point of obligation, we relied on four main values: conformity, benevolence, universalism and self-direction.
Commitment to the firm represents a focal point of obligation, one that is more prominent with employees high in the value of conformity. Schwartz (1992) described high conformity individuals as more restrained in actions, inclinations and behaviours that might violate social expectations or norms. They are more likely to inhibit their own feelings, conforming to rules and guidelines determined by society or their organisation. Indeed, Cullen et al. (2003) found a positive relationship between organisational commitment and individual perceptions of a principled ethical climate (i.e., one in which there is compliance with rules and regulations). In the same way, we can assume that employees who place a high value on conformity will feel a stronger obligation to the firm. They will thus seek the benefit of the company, even in the face of unethical requests. In the case of recruitment specialists, we expect that if they are high in conformity, they will be more vulnerable to conform even to unethical requests. Our first hypothesis can therefore be framed as follows: Hypothesis 1a: The more highly that recruitment specialists value conformity, the more likely they will be to comply with unethical requests from an organisational client.
A second focal point of obligation is the commitment to the welfare of society. This is especially pronounced in employees who highly value benevolence and universalism. Both values indicate the importance of preserving and enhancing the welfare of others (client or society) as the highest moral imperative. Schwartz (2012) noted that values of conformity and benevolence are similar, since they both enhance and promote cooperative and supportive social relations. Even so, they are motivated by slightly different sources. While conformity is informed by a desire to avoid negative results for the individual self, the motivational basis of benevolence incentivises cooperation.
Victor and Cullen (1988) suggested that an organisational benevolence climate encourages employees to focus on the benefits of others and to avoid unethical behaviours. Indeed, Fritzsche and Oz (2007) found a direct connection between altruistic values (composed of both universalism and benevolence values) and ethical work behaviours. That is, individuals high in universalism and benevolence tend to reject the idea of participating in unethical work behaviours. Similarly, we assume that recruitment specialists who value benevolence and universalism highly will try to benefit the welfare of others and minimise the potential damage from unethical demands. Therefore, they will feel less obligated to comply with unethical organisational requests. Our second hypothesis can therefore be framed as follows: Hypothesis 1b: The more highly recruitment specialists value benevolence and universalism, the less likely they will be to comply with unethical requests from an organisational client. Lastly, people motivated by self-direction are driven to control situations and search for autonomy to benefit their own desires. These employees do not feel obligated to conform to other standards but rely on their own interests. Indeed, Cullen et al. (2003) found 'egoistic' climates (i.e., self-interest and desire to maximise personal utility) in organisations to be negatively related to organisational commitment. They suggested that an egoistic firm climate signals to employees that the organisation is supportive of self-interested behaviours at the expense of the common good. Therefore, in an egoistic climate, normative values push individuals to make self-interested decisions rather than to consider the common good (Guerci et al., 2015). In relation to the egoistic climate, we assume that high selfdirection employees will try to maximise their utility and rely on self-interested decisions (Guerci et al., 2015;Martin and Cullen, 2006). Since recruitment specialists need to satisfy their organisational clients in order to receive their commission, we speculate that workers who rank self-interest highly will make efforts to comply with client requests -even unethical ones. Our third hypothesis can therefore be framed as follows: Hypothesis 1c: The more highly recruitment specialists value self-direction, the more likely they will be to comply with unethical requests from an organisational client. Previous studies have found that values exert a direct, as well as a moderating, effect on employee unethical behaviour (Barnett and Vaicys, 2000;Guerci et al., 2015). Thus, we can assume that a similar connection will also be found between those actions that represent 'silent' behaviours and intentions and those that represent more obvious behaviours. Let us take the example of an organisational client requesting that recruitment specialists send them only the résumés of younger candidates -an expression of non-merit-based ageism. The recruitment specialists are thus confronted with the dilemma of whether or not to pass along the résumés of suitable candidates aged 40 and above. If they follow the fi rst option, then another dilemma arises: do they inform the client that these candidates are over the requisite age limit or not? These two possible dilemmas represent different levels of behaviour activity on the part of the recruitment specialists. The fi rst is more of a silent non-compliance with the client's request (and is potentially non-confrontational, especially if the client does not notice the violation of their request), while the second represents more active, deliberate and potentially confrontational behaviour. In any case, the second dilemma extends as a consequence of the fi rst one. Thus, the focal point of an employee's ethical obligation will not only affect how they behave in each dilemma but will also moderate this connection. Therefore, our last hypothesis is as follows: Hypothesis 2: The employee focal point of ethical obligation moderates the connection between silent and deliberate unethical behaviour (conformity, benevolence/universalism and self-direction). Figure 2, below, summarises the second hypothesis such that high conformity employees will follow organisational rules. Thus, when pressed to decide how to respond to unethical requests, their reaction will be consistent with their behavioural predisposition. If they decide to comply with the original unethical request, they will maintain this compliance throughout the entirety of the client interaction. However, if they elect to not comply with the original unethical request, then they will consistently conform to the non-compliance behaviour. On the other hand, recruitment specialists high in self-direction are mostly obligated to their own self-interest, that is, they will try to maximise their own benefi t. As such, we expect them to alter their decision to satisfy their stakeholders from Phase 1 to Phase 2, based on moment-by-moment assessment grounded in the advantageousness of their position. Lastly, recruitment specialists high in benevolence/universalism will be more attuned to the moment-by-moment potential of causing offence to their stakeholders. Thus, they will change their behaviours based on conciliating the party they believe to be hurt in the fi rst phase. Consequently, recruitment specialists high in conformity will be consistent in their behaviours, while those high in self-direction and benevolence/universalism will alter their behaviours in different ways, depending on whether they are oriented toward the former or latter.

METHOD Participants
The participants in this study consisted of 176 HR recruitment employees (93% female, 7% male) working in an external employment and placement agency. The agency is part of a large, international company specialising in solutions in the field of HR, with more than 70 branches around the world. The Israeli agency has 12 branches with dozens of employees working in different specialisations, including high-tech, biotech, finance and banking, as well as recruiting for administrative, marketing, technical and low-tech positions. The Institutional Review Board approved the study protocol.

Measurement Dependent variable: complying with unethical requests
Based on previous studies (e.g., Fritzsche and Oz, 2007), we presented the participants with seven unethical dilemmas. In some, the unethical client requests were obvious (Scenarios 1, 2 and 7), while in others they were more ambiguous, but could have generated some negative bias against pregnant or disabled candidates (Scenarios 3, 4, 5 and 6). The requests were developed with three expert recruitment specialists, based on their experience. The different dilemmas appear in Table 1. After reading each dilemma, respondents were asked to answer two questions. In Phase 1, they were asked to rate their likelihood of passing along the candidate's résumé on a scale from 1-'will not send' to 9 -'absolutely will send'. A higher score translates into a greater intention to not comply with the client request. Since there were two different dilemmas types, we analysed the standardised mean difference (SMD) for the mean square of the two groups according to Cohen's d values. The results indicated that the differences are too small to be classed as significant between the two groups of ethical problems (d=0.06). The Cronbach's alpha for reliability test of all seven dilemmas was adequate (α=0.785). As such, we calculated a mean score for the seven dilemmas and named this variable 'ethic_ conform'. Higher results indicated greater participant willingness to reject unethical client requests, and thus greater reliance on ethical norms.
In the second phase, we measured willingness to disclose candidate discrepancy (e.g., lack of military service) to the client on a binary scale of 0 -'will not tell the client' or 1-'will tell the client'. The total score of participant admission to disclose the client request inconsistency was calculated as a new variable named 'ethic_mention'. Higher scores indicated greater employee conformity to organisational client requests. All participants were asked to answer the two questions/phases.

Independent variables Focal point of ethical obligation
The focal point of ethical obligation was measured based on the portrait values questionnaire (PVQ) (Schwartz et al., 2001). The PVQ contains short verbal portraits of different people. Each portrait describes a person's goal, desire or wish, according to the values in the sentence represented. The four values indicates different focal points of ethical obligation. Conformity represents employees more attuned to complying with authority (in this case, the firm, with its rules, regulations and client service). Benevolence and universalism represent employees more attuned to protecting the welfare of others. Finally, self-interest represents employees more attuned to complying with their own ambition, desire and independent thought. An example for conformity is as follows: 'She thinks people should follow rules at all times, even when no one is watching.' For benevolence, the following: 'He always wants to help the people who are close to him.' For universalism: 'She thinks it is important that every person in the world be treated equally. She believes everyone should have equal opportunities in life'. Finally, for self-direction: 'He thinks it's important to be interested in things'.
Respondents were asked to rank the degree to which each description resembles their own values on a scale ranging from 1 -'not like me at all' to 9 -'very much like me'. Each value was represented by several questions. We calculated a mean score for each value type. The reliability was sufficient for conformity (Cronbach's alpha=0.68), benevolence (Cronbach's alpha =0.85), universalism (Cronbach's alpha =0.89) and self-direction (Cronbach's alpha =0.78).

Control variables -demographic and other characteristics
Control variables included age ('age'), tenure ('tenure') and education ('education' 0-undergraduate degree, 1-postgraduate degree). Based on the small number of male respondents, we excluded 'gender' from our analysis. Table 1 summarises mean and standard deviation for the different dilemmas, as well as the percentage of recruitment specialists who agreed or disagreed to inform the client of candidate inconsistency. Table 1 reveals that, overall, recruitment specialists felt an obligation to the job applicant and would not conform to the unethical request. The highest resistance was registered in two requests: in the scenario with 1) a pregnant candidate and 2) a candidate undergoing fertility treatment (Scenarios 4 and 6). In contrast, resistance was lowest in two cases: 1) a client request for a secular candidate (while the applicant is extremely religious) and 2) a client request to screen out applicants with poor physical appearance. These two cases are linked by the fact that in both scenarios, the candidates' physical appearance cannot easily be concealed from the client.

RESULTS
In terms of the second question, a high percentage of respondents admitted willingness to disclose to the client their ethical inconsistency, if they passed along the applicant's résumé. Thus, the recruitment specialists who found the request unethical felt an obligation to inform the client of their infraction. Less agreement to inform the client was reported concerning candidate fertility treatment and military service. In both cases, this could be related to genderbased (mostly female) identification with the candidates (see Footnote 1). 1. In addition to professional requirements, the client asks that the candidate will be secular (not religious). You find a candidate that suits all the professional requests, who is ultra-orthodox.* 4.98 2.76 12.5 87.5 2. For a marketing job, the client asks that all candidates will be young and energetic. You find a perfect candidate, although he is over 40.
6.78 2.200 12.5 87.5 3. You are looking for a production worker for a factory. The candidate has extensive experience as a production worker. In the past year, he suffered an accident, which caused him a handicap. This limitation affects the output of his work.
6.06 2.26 9.1 90.9 4. The client requires interviewing the candidates before finalizing a decision on the position. You found a suitable candidate according to the professional requirements and after a face-to-face interview, you found out that the candidate is pregnant. 6. During a telephone interview, the candidate informs you that she is in the process of fertility treatment. As is well known, due to fertility treatments, the candidate may be absent more frequently than usual. 7. In addition to the professional requirement, the client asks that the candidate have served in the army. However, this work does not require military experience. You found a candidate that meets all the professional requirements, but did not serve in the army.
6.40 2.68 33.0 67.0 ¹ ethic_conform: the mean and standard deviation in regard to the question: "Will you send along the candidate resume to the client?" ². On a scale from 1-"will not send the resume to 9-"absolutely will send the resume". ³. Frequency in regard to the question: "Will you disclose the discrepancy to the client?" * All the scenarios ended with the question, "Will you send along the candidate resume to the client?" ' In this scenario, the question for HR employees was "If so, will you tell the candidate to alter his appearance before the interview with the client?"

Connection between focal point of ethical obligation, unethical evaluation and unethical disclosure
To assess our hypotheses, we conducted a hierarchical multiple regression test with ethic_conform as the dependent variable in the first model (Model 1) and ethic_mention in the second model (Model 2). In the first stage of each model, control variables were entered, while values were entered in the second stage. Lastly, readiness to pass along the résumé (ethic_conform) was entered in the third stage of Model 2. The results of both regression analyses are shown in Table 2, below.
The results of the first stage indicated that more educated recruitment specialists (those with more years of education, M.A. vs. B.A.) were willing to disregard the organisational client's requests and to send along the résumé. Neither age nor tenure were significant, indicating that ethical judgment is unaffected by experience in the employment agency. Entering the values in the second stage of Model 1 provided partial confirmation for the first hypotheses and indicated a significant contribution to the overall variance, F(9, 155)=2.842, p<0.01. However, of the four values, Notes: 1. Dependent variable: Ethic_conform -the readiness to send the candidate resume in spite of the unethical request; Ethic_mention -the sum of the seven decisions regarding whether the HR employee will inform the organizational client about the candidate mismatch with the clients' request. 2. Independent variables included -"benevolence", "self-direction", "universalism" and "conformity". 3. Standard errors presented in parentheses. 4. Coefficient significantly different from 0 at the 0.00% level ***, at the 1% level **, at the 5% level *.
only universalism was signifi cant (ß=0.252, t=2.397, p<0.05), confi rming H1b. The second model assessed the connection of the independent variables to readiness to mention inconsistency with client requests. In contrast to Model 1, age was positively connected to willingness to mention the candidate's inconsistency with the organisational client (ß=0.332, t=2.188, p<0.05). Older employees tended to admit they would point out client ethical inconsistency, if compelled to send along the candidate's résumé. Both benevolence and conformity values were signifi cantly connected, but in the opposite direction. Employees who highly appreciated benevolence were less willing to mention the inconsistency to the client (ß=-0.196, t=2.458, p<0.05), confi rming H1b. Conversely, high conformity employees were more willing to point out that the candidate was not compatible with client requests (ß= 0.154, t=2.658, p<0.05), confi rming H1a. Self-direction was not signifi cantly connected to willingness to disclose client inconsistency (ß=0.039, t=.538, n.s.), and as such, H1c was not supported.
Entering readiness to obey client request (ethic_conform) in the third stage contributed to the overall variance, adding a further 9.7%. Willingness to disregard unethical requests from the organisational client was negatively connected to readiness to inform the client about the candidate's problem (ß=-0.338, t=-4.433, p<0.001). This indicated that intention to behave was affected by employee evaluation that the request violated ethical standards, in addition to employee values and characteristics.

Moderating analysis
The last hypothesis suggests that the connection between evaluation of unethical requests and employee readiness to disclose to the client their incompatibility with ethical obligations will be moderated by employee values. A factorial analysis of variance (ANOVA) was conducted (see Footnote 2). This compared the main effect and interaction of the four values and the intention to send along the candidate's résumé on the readiness to disclose the discrepancy to the client. The results of the factorial ANOVA indicated a signifi cant main effect and interactions in three values: selfdirection, benevolence and conformity. For self-direction, there was a main effect for the readiness to send along the job applicant's résumé, F(1,172) = 10.318, p<0.001, and not for the importance of the value. A statistically signifi cant interaction was also found, F(1,172)=2.936, p<0.05. Figure 3 shows the mean differences between the two factors. The results show that high self-direction employees (i.e., conforming to their own agenda) felt obligated to both the candidate and the organisational client. As such, when they decided to send along the résumé (i.e., they disregarded the client's unethical request), they felt more of an obligation to inform the client about candidate discrepancy, as compared to employees who ranked self-direction low (M=0.69 and M=0.57, respectively). In contrast, self-directed employees who decided not to send along the résumé (i.e., they complied with the unethical request) were less likely to tell their organisational client about the candidate inconsistency, compared to employees who ranked selfdirection low (M= 0.80 and M=0.76, respectively).
For the benevolence value, both main effects were signifi cant: for readiness to send along the candidate's résumé (i.e., to disregard the client's unethical request), F(1,172)=13.56, p<0.001, and for the benevolence value, F(1,172)=3.782, p<0.05. The interaction was also signifi cant, F(1,172)= 8.488, p<0.01. Figure 4 shows the mean differences between the two factors. Employees who were high in benevolence (i.e., in obligation to the welfare of others) showed the same tendency to inform the client about the inconsistency with the requests (Cohen's d=0.13). Readiness to disclose the inconsistency to the client was more prominent when they agreed to send along the résumé and to not comply with the client's request (M=0.73 and M=0.50, for those who ranked the benevolence value high and low, respectively). The opposite direction was found when employees decided to comply with the client requests and not send along the résumé. In this case, high benevolence employees were less ready to inform the client about candidate inconsistency, compared to their low benevolence counterparts (M=0.81 and M=0.76, respectively).
In contrast to both self-direction and benevolence, conformity presented a reverse picture. The results showed a signifi cant main effect for sending along the candidate's résumé, F(1,172)=10.88, p<0.001. A signifi cant effect was also found for the interaction, F(1,172)=2.80, p<0.05. Figure 5 shows the mean differences between the two factors. For high-conformity recruitment specialists, those reporting that they would refrain from sending along the résumé (i.e., they would obey the client's unethical request) tended to disclose client inconsistency more than their lowconformity counterparts (M=0.83 and M=0.73, respectively), suggesting higher obligation to the client. On the other hand, those reporting that they would send along the résumé (i.e., disregarding the client's unethical request) exhibited lower tendency to disclose client inconsistency than their low-conformity counterparts (M =0.50 and M=0.65, respectively), suggesting higher obligation to job candidates -and not to the client.
Overall, the results supported our second hypothesis (see Figure 2). High-conformity employees were consistent in their decisions and behaviours for both dilemmas. On the other hand, employees high in self-direction and benevolence changed their decision between the fi rst and second dilemma.

Theoretical contribution
Previous studies have illustrated the role of ethical considerations in HR practices and functions (Fisher, 1999(Fisher, , 2005Fisher and Lovell, 2003;Winstanley and Woodall, 2000). However, it is less clear how ethical dilemmas affect external recruitment specialists. While HR employees are well aware of their obligation to the firm, recruitment specialists are caught between organisational clients who pay for their services and job-seeking candidates that are registered with the recruitment agency. Thus, the commitment of recruitment specialists is more nuanced and debatable and results in 'loyal matchmaker' dilemmas that affect their compliance with, or resistance to, unethical requests.
In agreement with Trevino's (1986) model, we found that ethical decision-making is both situational and affected by individual factors. In addition to ego strength, field dependence and locus of control, we proposed that personal values representing an individual's focal point of obligation (i.e., the moral source to which they pledge their strongest commitment) also impact on ethical decision-making. We found that recruitment specialists who prize the welfare of others (i.e., universalism) comply less with unethical client requests and would thus be less inclined to accede to the candidate's stipulations for the résumé. On the other hand, the found that other focal points of ethical obligation, including obligation to the firm (i.e., conformity value) were not connected to readiness to (not) comply with unethical requests.
As opposed to the first phase, both obligation to the firm (conformity) and welfare of others (benevolence) or themselves (self-direction) were found to influence employee decision-making to disclose client inconsistencybut in the opposite direction. Employees who ranked conformity highly also exhibited more consistency in their behaviours. In the case where they decided to go along with the unethical request and refrain from sending along the résumé, they also showed higher readiness to inform the client about the candidate's inconsistency. On the other hand, if they decided to send along the résumé, not acceding to the request, they were also less willing to inform the organisational client about the inconsistency. This behaviour can result from the tendency of high conformity individuals to avoid conflict (Schwartz, 2009). It also indicates that employees who are more obligated to the firm are more vulnerable to complying with manager or client requests, especially if the managers promote an unethical organisational culture. However, if the organisational culture encourages ethical behaviours, it is less likely that employees will obey unethical requests, whether implicit or explicit. The other two values of self-direction and benevolence showed different patterns of behaviour compared to highconformity employees. While the latter group was consistent with their behaviours, employees high in self-direction and benevolence compensated their first decision with the second one. Those who disobeyed the unethical requests in the first request showed a higher tendency to inform the organisational clients about the inconsistency. For the employees who were obligated to their self-interest (i.e., self-direction value), this behaviour can be explained either by an instinct to preserve their independence (Schwartz, 2009), resulting in a refusal to take a clear stance, or by serving their own utilitarian ends by satisfying all sides. Those who perceived that a stakeholder was offended by their initial decision would subsequently alter their behaviour, in an attempt to placate any injured party. In doing so, they could theoretically avoid alienating any particular party and at the same time preserve their own interests.
The employees high in benevolence showed a similar pattern of behaviour, which was somewhat surprising. Indeed, Schwartz (2009) defined people who are high in benevolence as being more attuned to the welfare of others, which explains why these employees felt less of an obligation to disclose client inconsistency to begin with. However, those who complied with the client tended to change their decision, resulting in unethical behaviour. This behaviour is similar to self-interested employees but results from a different motivation. Since high-benevolence individuals feel obligated to preserve the welfare of others, they are inclined to compensate those who were offended by past decisions. Thus, high-benevolence recruitment specialists make compensatory gestures to their organisational client, if they feel that the client had been damaged by their initial decision. The desire to preserve the welfare of others led them to alter their behaviours based on who they felt were mostly hurt from their previous decision. This behaviour should be further explored in future studies, that is: How do perceived obligations to others affect unethical decision-making?
The results of the study indicate that human resource workers, including recruitment specialists, are not necessarily driven by social motivation. Some may see their role as being more attuned to the welfare of others. However, others see themselves as handmaidens of efficiency, strategic partners, managers' consultants and so on. We suggest that HR workers are differentiated by their basic motivations, as manifested in their values. While some may feel more obligation to the welfare of society (resulting from a high-benevolence value and not because they are working in HR or recruitment agencies), others, in the same company, can be motivated by values of self-direction. Recruitment agencies do not represent the candidate; rather, they act as agents and intermediaries between the candidate and the organisation. However, since they serve this intermediary role, they also become 'gatekeepers', which can lead to ethical issues. The way recruitment workers resolve these ethical issues indicates how they describe their obligations -to the firm or to the candidate. However, this does not mean that the recruitment worker will always take the side of the candidates and feel obligated to represent them, but that the order of the recruitment specialist's values impacts on how they resolve the dilemma.

Implications for practice
In this study, we noted the loyal matchmaker dilemma of recruitment specialists, that is, to whom should the recruitment specialist show commitment? On the one hand, external recruitment agencies need to represent the interests of their client organisations, especially the HR departments. After all, these are the agencies' paying customers. On the other hand, they are also obligated to the job candidates, whom they send to the organisation wishing to fill open positions. This triangulation of interests has informed several lawsuits and led to legislation, especially in the form of regulations for hiring temporary employees (Wears and Fisher, 2012). Recruitment agencies must clearly articulate that they have a responsibility both to job candidates and to organisational clients. This can be achieved through communication, ethical codes, regulation and the development of an ethical climate in the recruitment agency.

Communication of ethical dilemmas and ethical standards
Communication plays a central role in raising awareness about ethics and ethical standards that are expected by the agency (Jha et al., 2017). Recruitment agencies should allow constant dialogue regarding the ethical commitment and obligation of the firm to both clients and candidates. The agency must highlight the problem of triangulation of interests that could cause a clash of commitments. Managers should encourage workers to raise concerns and speak about ethical inconsistencies or unethical requests from the clients. Employees should feel that they can consult on cases that are unethical or likely to harm candidates. As Fisher (1999) discovered, HR specialists are more attuned to taking complex or sophisticated ethical stances more frequently, compared to accountants and financial co-workers. Therefore, they must find ways to deliberate on ethical dilemmas when they face them. This can be achieved through ethical meetings or through a formal mechanism such as the appointment of a compliance officer to advise employees and managers on adopting appropriate behaviours. Another option is to provide the agency with an ethical decision-making support system, such as a web-based decision support system (DSS), which will assist and guide recruitment specialists who are experiencing ethical dilemmas (Carbery et al., 2014). All these methods should translate into an atmosphere in which dilemmas can be freely spoken of and deliberated on.

Legislation and ethical codes
To ensure compliance with ethical standards, recruitment agencies should provide clear codes and regulations for ethical behaviours. The agency can provide codes based on governmental rules and regulations, as well as professional codes (Fisher and Lovell, 2003;Manroop et al., 2014). Since recruitment agencies sometimes serve as joint employers with their organisational clients, there are specific rules that should be taken into consideration and these should be clearly transmitted to employees (Seipel, 2017).
Even so, our results indicated that employees differ in their focal point of obligation, and this affects how they treat firms' regulations. In this sense, ethical codes should also be formulated, implanted and enforced, based on employees' focal point of obligation -customised to fit a preference for the company, one's own self-interest or welfare of others. For instance, high-conformity employees are more attuned to complying with the firm and its regulations. As such, they need explicit ethical codes and regulations. High-benevolence employees feel a stronger obligation to preserve the welfare of others. As they are focused on the welfare of others, they might need to be explicitly informed by their managers of how their behaviours might affect external stakeholders such as clients, suppliers or even competitors. In addition, the ethical codes should be accompanied by training conformity to ethical rules and codes (Manroop et al., 2014). Training on reactions towards unethical demand based on ethical codes will lead to more reliable and predictable reactions, as well as reducing the hazardousness of unethical or illegal behaviours.

Limitations and suggestions for future research
As with most studies, this research has several limitations that should be acknowledged. As noted, the focus herein was on ethical decision-making. To measure unethical behaviour, we used ethical dilemmas. While this method is common in studies on ethics (e.g., Finegan, 1994), there is a difference between intention to behave and actual behaviour. As such, future studies should evaluate the connection between ethical intentions and actual behaviour in work situations.
Another limitation results from focusing on recruitment specialists working at an external recruitment agency. We noted that this group is more vulnerable to the loyal matchmaker dilemma which pressures them to complying with unethical behaviour (Seipel, 2017). However, this uniqueness also makes it difficult to generalise results to other ethical organisational situations. Future studies should extend this research into effects of personal values on ethical behaviour in other organisational dilemmas, such as their effect on manager decision-making and performance measurement (Fisher and Downes, 2008;Watson, 2003).
Lastly, we explored the tendency to conform to unethical client requests, based on Schwartz's basic human values theory (Schwartz, 1992(Schwartz, , 2012. Therefore, we focused on the individual level, distinct from previous research that has tended to privilege ethical climates in an organisational setting (e.g., Guerci et al., 2015;Parboteeah et al., 2010). Few studies have applied Schwarz's model to measure ethical decision-making, and so this research should be complemented with further iterations using this method. More studies are needed to reveal the effect of personal values on ethical judgment, as well as the effect of organisational ethical culture and its effect on readiness to comply with unethical behaviours based on employee inner values.
In conclusion, recruitment specialists are caught between a rock and a hard place. This study shows that the loyalty of recruitment specialists depends on more than a single factor and that their personal values, enmeshed with the particulars of a given situation, are determinative. As the global workplace continues to change and the demand for flexibility in hiring increases, the reliance on external agencies will increase, bringing with it the problem of the loyal matchmaker.

Footnotes
1. In Israel, military service is mandatory for Jewish men, but not women, if they are religious. This request might evoke a perception of discrimination against non-military civilians, especially religious job seekers. 2. Dummy variables for values and readiness to send the résumé were first created based on the median score of variables.