Who asks and who receives in salary negotiation

Abstract

The influence of individual differences and negotiation strategies on starting salary outcomes was investigated. A sample of 149 newly hired employees in various industry settings participated in this study. Results indicated that those who chose to negotiate increased their starting salaries by an average of $5000. Individuals who negotiated by using competing and collaborating strategies, characterized by an open discussion of one's positions, issues, and perspectives, further increased their salaries as compared to those who used compromising and accommodating strategies. Individual differences, including risk-aversion and integrative attitudes, played a significant role in predicting whether or not individuals negotiated, and if so, what strategies they used. Copyright © 2009 John Wiley & Sons, Ltd.

Introduction

Starting salary inequities can have a compounding effect on individuals' career earnings. For example, assuming a 5 per cent pay increase each year over a 40 year career, a twenty five year old employee that starts at $50 000 would earn $634 198 less than an employee starting at $55 000 by the time s/he reached the age of 65. In addition to the financial implications of a lower starting salary, the perceptual outcomes of the salary negotiation process are meaningful. The justice literature indicates that receiving less than satisfactory outcomes reduces perceptions of fairness (Colquitt, 2001; Lind & Tyler, 1988), which produce beliefs that one is not being compensated appropriately for contributions they bring to an organization (Porter, Conlon, & Barber, 2004).

To date, limited field research on salary negotiation has found that those who choose to negotiate generally increase their salaries. The few studies in this area have examined the effect of gender on salary negotiation outcomes, finding that women are less likely to negotiate (Babcock, Gelfand, Small, & Stayn, 2006; O'Shea & Bush, 2002) and that men receive larger salary negotiation payoffs (Gerhart & Rynes, 1991). A more detailed study of individual differences and choice of negotiation strategy, looking beyond gender effects, should give us a broader understanding of why certain individuals reap more benefits than others in salary negotiation. Despite a call to “look further at how individual differences are related to process features that determine bargaining outcomes” (Barry & Friedman, 1998: p. 357), we know little about how individual differences and choice of negotiation strategies affect important outcomes for new employees.

There is a large body of research on general negotiation practices supporting the notion that effective negotiation can yield better outcomes for involved parties (Lewicki, Barry, Saunders, & Minton, 2004). Findings have been drawn primarily from laboratory research, through the use of simulated bargaining scenarios and role-play methodology. The strength of this body of controlled experimental research lies in the precise manipulation of a small number of factors. Yet there have been frequent calls from negotiation researchers to test their findings in field settings. Recent studies (e.g., Van Kleef, De Dreu & Manstead, 2004) highlight the importance of emotion in negotiation and raise questions about the ability of simulations to duplicate the full emotional tensions and ego involvement that are present in real world negotiations with tangible outcomes, such as salary, at stake.

The few studies (e.g., Babcock et al., 2006; Gerhart & Rynes, 1991; O'Shea & Bush, 2002) on salary negotiation, examine the link between structural factors and gender with one's decision to negotiate and hard salary outcomes. Gerhart and Rynes (1991) provide the first empirical study of evidence that propensity to negotiate leads to a salary payoff. They surveyed graduating MBA students to examine structural (i.e., number of alternative job offers, attractiveness of initial offer) and individual (i.e., gender) determinants of negotiation propensity (e.g., whether or not an individual chose to negotiate) and bargaining payoff (i.e., starting salary). They found that negotiations were more likely when initial salary offers were low, and when the number and attractiveness of alternative job offers was high. Gender differences did not significantly predict propensity to negotiate, but among those who negotiated, men negotiated larger payoffs.

O'Shea and Bush (2002) surveyed recent college graduates on a variety of structural factors that affect salary negotiation outcomes. They found that applicants given the option to present their salary needs were more likely to negotiate than those who were not, and those who had prior work experience were more likely to negotiate than those who did not. Consistent with Gerhart and Rynes' findings, they found that women were no less likely to negotiate, yet in contrast to Gerhart and Rynes, women were as successful as men in salary negotiations. Neither O'Shea and Bush (2002) nor Gerhart and Rynes (1991) focused directly on strategies used during negotiations. Instead they compared structural factors (e.g., alternative offers) on decisions to negotiate and financial outcomes of negotiation. While this is an important first step, it takes a “black box” approach, linking contextual inputs to negotiation outcomes without examining the role of the negotiation process itself. Further, early studies examined only the reported negotiated salaries, without consideration of the role of perceptions of fairness and satisfaction with negotiation process. Perceptions of fairness during the negotiation process may play a pivotal role in forming job attitudes such as job satisfaction and commitment, which affect job behaviors such as performance and turnover.

The purpose of this study is to propose and test an input-process-outcome model of the salary negotiations process using two field samples of salary negotiations from business and academia. We examine relationships between individual differences, job negotiation strategies, and salary negotiation outcomes (increased starting salary, reactions to job offers). Figure 1 illustrates the model we examine in this study. Our primary thesis is that individual differences influence one's choice of whether and how negotiation strategies are employed, above and beyond contextual factors. We also argue that the type of negotiation strategy employed predicts important salary negotiation outcomes, including salary increases and individuals' reactions to negotiated outcomes.

Figure 1. Proposed I-P-O model of salary negotiations

Negotiation Strategies

The negotiation process is the interaction that occurs between parties before a final outcome is reached (Thompson, 1990). We refer to a job applicant's negotiation strategy as a broad, general plan of action used to attain the negotiator's goals. Strategies differ from tactics on a matter of scale. Tactics are “short term, adaptive moves designed to enact or pursue broader (or higher level) strategies…” (p. 92; Lewicki, Saunders, & Minton, 1997), whereas strategies include the use of multiple tactics and represent stylistic approaches to the negotiation process.

We examine five negotiation strategies that describe approaches to managing conflict: Collaborating, competing, avoiding, accommodating, and compromising (Dallinger & Hample, 1995; De Dreu, Evers, Beersma, Kluwer, & Nauta, 2001; Rahim, 1983). These five strategies have been frequently described in the literature and are based on the dual-concern model (Blake & Mouton, 1964; Rahim & Bonoma, 1979). Collaborating, accommodating, and compromising strategies all derive from a high concern for others or a pro-social orientation (Blake & Mouton, 1964; De Dreu, Weingart, & Kwon, 2000; Pruitt & Rubin, 1986; Rahim & Bonoma, 1979). This orientation lends itself to a cooperative, integrative approach that enables a negotiator to focus on helping others achieve their goals and find mutually beneficial solutions. Competing conflict management strategies stem from self-concern, or an egoistic orientation (Blake & Mouton, 1964; De Dreu et al., 2000; Rahim & Bonoma, 1979). Self-concerns focus on doing what is perceived as best for oneself, such as winning or reaching the best self-outcome. Each of the five negotiation strategies are described in more detail below.

Collaborating

This strategy has also been referred to as integrating or problem solving. Relative to the dual concerns model, this strategy represents a high concern for attaining one's own outcomes as well as a high concern for whether the other party attains their desired outcomes. A collaborative strategy is represented by a desire to exchange meaningful and accurate information in order to reach an agreement that is best for all parties involved. There is an emphasis on discovery of the basic interests of those involved in the negotiation, in order to craft a solution that meets both parties' interests. This strategy is especially appropriate when there is value to synthesizing ideas to develop better solutions, when time is available for negotiation, and when both parties have an investment in the outcomes (Lewicki et al., 2004).

Competing

Also called contending or dominating, the competing strategy represents a greater concern for one's own outcomes and a lower concern for other's outcomes. Tactics utilized in a competing strategy can include persuading, threatening, misrepresenting, and asserting.

Accommodating

Also referred to as obliging or yielding, an accommodating strategy to negotiation represents high concern for others and low concerns about one's own outcomes. Negotiators pursuing an accommodating strategy are more interested in having others attain their desired outcomes. While this strategy has disadvantages when one is trying to reach agreement on issues that are important, this strategy may be appropriate in situations where one's focus is on the longer term relationship or one is negotiating from a position of limited power.

Compromising

Compromising involves some level of concern for one's own outcomes and some level for others' outcomes. The use of a compromise strategy would include the use of a give-and-take approach with a desire to reach an acceptable middle ground.

Avoiding

This approach involves dodging situations that would involve negotiation. While there are situations where an “avoid” approach would be effective, often in salary negotiations an avoid approach “leaves money on the table.” The dual concerns model considers the avoiding strategy as one that represents low concern about one's own and other's outcomes.

Risk aversion and its impact on choice of negotiation strategy

We examined risk-aversion for its particular importance in the context of a job negotiation. Unlike other more general bargaining scenarios, salary negotiation is optional, and there is at least some level of risk to entering into negotiation once an offer has been put forth. Applicants could accept the current offer without negotiation to secure the job. Further, individuals who are risk-averse may be more inclined to frame the negotiation process as a win–lose process. Risk averse individuals may fear the potential damage that it does to ongoing relationships. By electing to enter into negotiation, they are, in effect, taking a risk of losing the opportunity by signaling that the current job offer is not acceptable.

Individual risk taking behavior has both state and trait components. One premise of prospect theory (Kahneman & Tversky, 1979) is that risk-taking behavior is inconsistent across situations—a person will take risks under some conditions and not in others. However, there is a body of research that appears to contradict prospect theory, showing that decisions about when to take risks are strongly rooted in personality (Nicholson, Soane, Fenton-O'Creevy, & Willman, 2005). Bottom and Studt (1993) Argue that the manner in which individuals frame a negotiation are rooted in one's perception of risk. The more risk one sees in the process of negotiation, the less assertive methods one would be willing to use in negotiations. Risk averse people may not be willing to chance the destabilization of relationships with aggressive negotiation strategies, because the impact on future relationships is unknown.

We expect that individuals who are more risk averse will frame a salary negotiation as more of a risk and thus be less willing to enter into negotiation. When risk averse individuals opt to negotiate, they will likely take a more passive approach. Risk averse people will be less likely to engage in more active and aggressive negotiation behavior. Competitive and collaborative approaches to negotiation require a striving for a different outcome. This different outcome may have the potential of a higher payoff, but also requires engagement in a negotiation process that may be perceived as potentially causing damage to the relationship, creating a less stable social dynamic.

H1a: Personality-based risk aversion will predict whether or not individuals choose to negotiate, above and beyond the effects of contextual power. Individuals who are high in risk aversion will be less likely to negotiate than individuals lower in risk aversion.

H1b: Among individuals who chose to negotiate, personality-based risk aversion will explain how individuals choose to negotiate, above and beyond the effects of contextual power. Individuals who are high in risk aversion will be more likely to use an accommodating strategy than individuals lower in risk aversion.

H1c: Among individuals who chose to negotiate, personality-based risk aversion will explain how individuals choose to negotiate, above and beyond the effects of contextual power. Individuals who are high in risk aversion will be less likely to use a collaborating strategy than individuals lower in risk aversion.

H1d: Among individuals who chose to negotiate, personality-based risk aversion will explain how individuals choose to negotiate, above and beyond the effects of contextual power. Individuals who are high in risk aversion will be less likely to use a competing strategy than individuals lower in risk aversion.

Attitudes

One attitudinal individual difference that is likely to play a role in predicting choice of negotiation strategy is to what extent one holds an integrative view of the negotiation process. In general, people with integrative perspectives on negotiation believe that information should flow freely during negotiations and effort to understand issues on both sides. Individuals who hold integrative attitudes believe there should be an emphasis placed on commonalities and a search for solutions that meet the needs of both sides. Galinsky, Maddux, Gilin, and White (2008) found that the ability to take others' perspectives in a negotiation context increased their ability to find unobvious solutions that were mutually beneficial to both negotiating parties.

In addition to considering others' perspectives, integrative attitudes underlie a belief that both parties are likely to benefit from working together and a belief that a trusting, cooperative approach can lead to the best solutions for all parties. Beersma and De Dreu (1999) found that people with pro-social motives, meaning those who seek good outcomes for themselves and others, were able to reach more integrative negotiation agreements. Social motives may, in part, stem from integrative attitudes that convey the value of finding win–win solutions to a negotiation. People with integrative attitudes believe that for ongoing relationships, a jointly obtained solution is more likely to set up a base for cooperative future interaction than win–lose solutions. In contrast, individuals with a less integrative perspective view negotiation as a fixed-sum process, and are more likely to maximize personal gain and minimize losses with “win–lose” orientations. Individuals that value outcomes and relationships and view negotiation as an integrative vehicle to reach agreements are more likely to negotiate salaries and to use collaborative strategies to reach agreement.

H2a: Integrative attitudes will predict whether or not individuals choose to negotiate, above and beyond the effects of contextual power. Individuals who hold more integrative attitudes about negotiation will be more likely to negotiate than individuals with less integrative attitudes about negotiation. H2b: Among individuals who chose to negotiate, integrative attitudes will explain how individuals choose to negotiate, above and beyond the effects of contextual power. Negotiation Strategies and Outcomes.

Laboratory research has demonstrated that the type of negotiation strategy used has a dramatic impact on the process and outcome of the negotiation (Pruit, 1983; Thompson & Hrebec, 1996). Negotiation research typically differentiates between competitive (forcing) and cooperative (problem-solving) approaches. Theory and empirical studies have found that cooperative, problem-solving approaches rather than forcing approaches lead to more integrative (win–win) solutions (Lewicki et al., 1997; Pruit & Carnevale, 1993). Problem solving behaviors indicative of cooperative approach involve the exchange of information, preferences and priorities; whereas competitive approaches use tactics like persuasion, claiming, and the use of threats. It is unclear to what extent experimental research generalizes to real world salary negotiation contexts, where more variables are necessarily operating simultaneously.

Competing and collaborating are both active strategies that work to maximize the outcomes of the negotiator. Whereas collaborating involves a dual focus on maximizing joint outcomes, competing focuses more directly on achieving an individual's goals. We expected that these two negotiation strategies will be linked with actual increases in negotiated salary because both strategies include behaviors and tactics that lead to an overt expression of one's interests and presentation of information towards the goal of increasing one's salary.

H3a: Among individuals who chose to negotiate, those who use a collaborating strategy will increase their salary offer.

H3b: Among individuals who chose to negotiate, those who use a competing strategy will increase their salary offer.

Pay dissatisfaction has been linked with undesirable employee outcomes and behaviors, such as performance decrements, lateness, and job seeking (Heneman & Judge, 2000; Williams, McDaniel, & Nguyen, 2006). First, social exchange theory posits that people react more favorably when the outcomes they receive are positive (Homans, 1961). Further, we expect that strategies that involve a dual focus on achieving one's interests and creating or sustaining long term relationships with the other parties involved in the negotiation will create a more balanced negotiation process, and that negotiators will reflect back on their negotiations as having been fair and having been satisfied with the negotiation process and outcome. The justice literature lends support to this argument. Namely, Adams (1965) equity theory maintains that attitudes (e.g., fairness, satisfaction) will be adversely affected by perceived inequities in outcomes. Furthermore, according to the work of Thibaut and Walker (1975) organizational procedures are more likely to be perceived as fair when individuals are given the opportunity to influence the decision process; what is known as voice (Gilliland, 1994). Negotiation strategies that involve a dual focus give the negotiator “voice” in the negotiation process, in that they have the opportunity to influence their outcomes. Strategies such as accommodating, in which negotiators give up their voice, will likely result in overall lowered perceptions of fairness and lower levels of satisfaction.

Within our I-P-O framework, we also hypothesize that negotiating strategies are the mechanism through which personality based risk aversion and negotiation attitudes influence salary negotiation outcomes.

H4a: Among individuals who chose to negotiate, those who use a collaborating strategy will have more positive reactions toward their negotiated outcomes.

H4b: Among individuals who chose to negotiate, those who use a compromising strategy will have more positive reactions toward their negotiated outcomes.

H4c: Among individuals who chose to negotiate, those who use an accommodating strategy will have more negative reactions toward their negotiated outcomes.

H5: Negotiation strategies will mediate relationships between risk aversion and attitudes and negotiation outcomes.

Method

Participants

A sample of 149 newly hired employees in various industry settings participated in the present study. More specifically, data were gathered from two primary sources; tenure-lined faculty new hires for a large Mid-Atlantic University (n = 71) and from new hires in local industry in the same region (n = 78). Names and email address of all faculty hired within the previous three years were obtained from the University Human Resources office. Similarly, part-time MBA students enrolled in classes from the same University were recruited for participation this study. Faculty new hires and MBA students, who had recently participated in a job search process, received an email with information concerning the study goals and procedures. Those who were interested in participating were directed to a website containing survey materials. Only individuals who had been hired within the last three years were recruited for participation. This ensured that our participants were not too far removed from the negotiation process. The sample was comprised of 52 (34 per cent) women and 91 (59 per cent) men; six individuals did not report gender information. The sample was primarily Caucasian (77 per cent). To further protect anonymity of respondents, participants did not report their actual age, rather reported the age band that best reflected their age at the time of survey completion. Thirty-one (20 per cent) participants fell in the 20–29 age range, 55 (36 per cent) participants were in the 30–39 age range, 30 (20 per cent) were 40–49 years of age, 25 (16 per cent) were 50–59 years of age, and only 1 (0.6 per cent) participant was 60 or older. Seven participants did not provide age data. On average respondents had participated in 2.65 previous job negotiation processes and 73 per cent had negotiated at least once in the past. In the current process, 110 (74 per cent) participants chose to negotiate while 39 (26 per cent) did not negotiate the terms of their offer (in other words, they avoided the negotiation process). Salary gains were reported only for those individuals who negotiated. Of those who did negotiate, they reported average salary gains as a result of negotiation of $4913 (σ = $4581).

Procedure

Participants first completed a brief section assessing individual factors. At the end of this section, participants were asked to indicate whether or not they negotiated for their current position. Those who indicated “no” were directed to a section assessing reasons for not negotiating and other demographic information. Those who indicated “yes” were directed to a section containing items assessing their negotiation strategies, outcomes, and other demographic information.

Prior to testing hypotheses, a variable was created, labeled “sample” (coded “1” for industry and “2” for academia), representing whether the respondent was negotiating for an industry or academic position. Bivariate correlations were run to determine whether sample was related to important study variables. Results determined that academic participants were more likely to negotiate than industry participants (r = 0.32, p < 0.05). Additionally we should point out the differences regarding the average salaries raised within each group. Namely, industry participants raised their salary by an average of $5939 (σ = $5406), while academic negotiators raised their salary by an average of $4123 (σ = $3696), this difference was marginally significant (t = 1.84, p = 0.07). Because of these differences we control for sample during hypothesis testing. Additionally, research suggests significant gender differences in negotiation success (Gerhart & Rynes, 1991). While this research has met with mixed results (Gerhart & Rynes, 1991; O'Shea & Bush, 2002), in our study gender was significantly related to salary raised (r = −0.23, p < 0.05) such that men generally negotiated for higher starting salaries. We therefore also control for gender during hypothesis testing.

Measures

Integrative attitudes

A three-item scale was developed to assess whether participants demonstrated integrative attitudes toward negotiations (with higher values reflecting more integrative attitudes about negotiation). An example item from this scale is: “Negotiation is almost always a process of one's gain over the other” (those who responded affirmatively to this question would be demonstrating a less integrative attitude). All scales were measured on a five-point scale ranging from “1” (strongly disagree) to “5” (strongly agree). The reliability for this scale was α = 0.70. Full items from this scale, along with all study scales, are available in Appendix A.

Personality-based risk aversion

We assessed risk aversion using an adapted version of Slovic's (1972) Risk Aversion Scale. An example from this four-item scale is: “I am not willing to take risks when choosing a job or a company to work for.” Coefficient α for this scale was, α = 61. Items were measured on a five-point scale ranging from “1” (strongly disagree) to “5” (strongly agree).

Negotiation strategy

Avoidance or the decision to negotiate was assessed with a single dichotomous item, “After receiving a job offer, did you negotiate the terms of the offer?” We used this item as a proxy for the “avoid” negotiation strategy. Individuals who chose not to negotiate for whatever reason were considered as avoiding negotiation, whereas individuals who negotiated utilized strategies other than avoid. The remainder of the negotiation strategies was assessed using both an adapted version of Rahim's (1983) Negotiation Scale and items created by the researchers for purposes of this study. A seven-item competitive strategy scale developed by the researchers was included to assess the degree to which participants adopted a competitive strategy during job negotiations. An example from this scale is: “During negotiation, I tried to persuade the organization to better my offer by threatening to withdraw from the process” (α = 0.62). The remaining three scales were adapted from Rahim (1983) and included items assessing the extent to which participants' strategies could be classified as: Collaborative (“I exchanged accurate information with the organization to come to a joint agreement,” seven items, α = 0.79), accommodating (“I felt myself trying to accommodate to the wishes of the organization,” four items, α = 0.72), and compromising (“I used ‘give and take’ so that compromise could be made,” three items, α = 0.63). Items were measured on a five-point scale ranging from “1” (strongly disagree) to “5” (strongly agree).

Outcomes

Two primary outcomes were assessed in the present study. First we asked respondents who negotiated to tell us how much they were able to raise their starting salary in dollars. Second, we created a three-item measure labeled Reactions to Negotiations to assess applicants' affective reactions to the negotiation process. Participants were asked to recall how satisfied they were with their final offer at the time of hiring (“At the time I signed the offer, I felt my final negotiated job offer package, salary and other job-related benefits, was fair and reasonable”), how satisfied they are now with their agreed upon job offer (“Looking back on it, I am still satisfied with the offer I received”), and their perception of the fairness of the negotiation process (“I felt the negotiation process was fair and reasonable”). The coefficient α reliability of this scale was α = 0.84 and items were measured on a five-point scale ranging from “1” (strongly disagree) to “5” (strongly agree).

Power

Power was used as a covariate in this study representing the negotiating context, and was operationalized as informational power, additional job offers, and experience. A three-item market-information scale was created by the researchers to assess how much information participants gathered about the organization with which they were negotiating. An example item from this scale is: “How much information did you gather about previous offers made by the organization?” Items were measured on a five-point scale ranging from “1” (nothing/none) to “5” (very much). Alpha for this scale was 0.70. Participants were asked to indicate the number of alternate job offers they had received at the time. Finally, a three-item experience scale was created by the researchers to assess negotiators previous negotiation and job experience. An example item included: “Prior to this job offer, how many times had you negotiated job offers?” Alpha for this scale was 0.79.

Demographics

Questions assessing age, race, and gender were asked in this section.

Descriptives and CFA

Table 1 presents means, standard deviations, and correlations among all study variables. A confirmatory factor analysis was run to provide evidence for the construct validity of the individual differences (integrative attitudes and risk aversion) and strategy variables (competing, collaborating, compromising, and accommodating) examined in the present study. Hu and Bentler (1999) suggest multiple fit indices be used for judging model fit. SRMSR values of <0.08 are considered a “relatively good fit for the model” and values <0.10 are considered as “fair” (Browne & Cudeck, 1993). Additionally, Browne and Cudeck (1993) suggest reporting root mean square error of approximation (RMSEA) values; with values 0.08–0.06 representing acceptable fit and values <0.06 representing good fit. As presented in Table 2, the best fitting model was one in which items loaded onto their respective factor (i.e., integrative attitude items loading onto attitudes factor; competing items loading onto competing scale). Any deviation from this model resulted in poorer fit.

Table 1. Scale mean, standard deviations, and correlations

		Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14
Significant correlations at p < 0.05 in bold. Avoid coded such that 0 = did not negotiate, 1 = negotiated; Sample coded such that 1 = industry, 2 = academic; Gender coded such that 1 = male, 2 = female. Note: Correlations with strategy variables and negotiation outcomes only include those individuals who engaged in salary negotiations, N = 110. Remaining correlations include full sample, N = 149.
1	Salary raise	4913	4581	–
2	Reactions	3.87	0.83	282: 0.25	0.84
3	Avoid strategy	0.74	0.44	–	–	–
4	Competitive strategy	3.16	0.55	0.39	0.24	–	0.62
5	Collaborative strategy	3.96	0.50	0.24	0.51	–	0.41	0.79
6	Accommodating strategy	2.82	0.76	−0.06	−0.44	–	−0.32	−0.36	0.72
7	Compromising strategy	3.44	0.64	0.16	0.32	–	0.25	0.55	−0.15	0.63
8	Risk aversion	2.41	0.61	−0.04	−0.10	−0.20	−0.31	−0.40	0.25	−0.22	0.61
9	Integrative attitude	3.56	0.78	0.22	0.26	0.23	0.22	0.31	−0.14	0.16	−0.17	0.70
10	Market information	2.96	0.93	0.14	0.17	0.08	0.30	0.38	−0.15	0.30	−0.08	0.16	0.70
11	Offers	1.81	0.78	0.10	0.17	0.18	0.02	−0.06	−0.08	0.05	−0.03	0.08	0.21	–
12	Experience	3.27	1.12	0.34	0.28	0.06	0.31	0.20	−0.26	0.05	−0.21	0.07	0.07	0.14	0.79
13	Sample	1.36	0.48	−0.20	−0.25	0.32	−0.21	−0.06	0.05	−0.14	−0.02	0.10	−0.07	0.02	−0.03	–
14	Gender	1.47	0.50	−0.23	−0.10	0.00	−0.03	−0.18	0.04	−0.03	−0.09	0.18	−0.12	−0.10	−0.18	0.03	–

Table 2. Confirmatory factor analysis of key study variables

Model	X^2	df	Δχ^2	SRMR	CFI	RMSEA
Model 1 specified six distinct factors in which items loaded onto their respective factors (integrative attitudes, risk aversion (the two IV's); and competing, collaborating, compromising, accommodating (the four strategies)). Model 2 specified four factors; namely two distinct IV's (integrative attitudes and risk aversion) and two strategies (self-focus vs. others-focus). Model 3 specified five factors; namely one IV (all independent variable items loaded onto one variable representing IV) and four distinct strategies. Model 4 specified three factors; namely one IV (all independent variable items loaded onto one variable representing IV) and two strategies (self-focus vs. others-focus).
1	423.58	335	—	0.09	0.87	0.04
2	529.05	344	105.47**	0.10	0.73	0.08
3	479.85	340	56.27**	0.11	0.80	0.06
4	583.92	347	160.34**	0.11	0.66	0.09

As we discuss above, our methodology relied on self-report data collected at a single administration. Such research designs can be susceptible to common method variance (Campbell & Fiske, 1959). We conducted Harman's one-factor test to determine the extent to which common method variance was driving underlying relationships between study variables. Podsakoff and Organ (1986) note the Harman one-factor test is among the first statistical procedures to test and control for common method variance, and it continues to be a commonly used analysis (Ambrose & Mitchell, 2007; Holtz & Harold, 2008; Podsakoff, Todor, Grover, & Huber, 1984). According to Harman's one-factor test, considerable common method variance is evidenced if either a single factor emerges in a factor analysis or the first factor extracted accounts for a considerable amount of the variance in items included in the factor analysis (Podsakoff & Organ, 1986). Results of the Harman one-factor test revealed that neither of the aforementioned conditions was met, as expected multiple factors were extracted and the first factor accounted for only 19.43 per cent of the total variance in our variables. Thus, a dominant general factor is not accounting for the majority of the variance in individuals' survey responses.

Test of hypotheses

To examine our hypotheses, we focused on two main criteria (outputs): Reported gain in salary due to negotiation (in dollars) and reactions to the negotiation process. Processes in our I-P-O model focused on negotiation strategies. First, did respondents decide to negotiate? For those that did negotiate, further information was gathered on the strategies negotiators used, namely competing, collaborating, compromising, or accommodating strategies. Finally, inputs consisted of contextual power, and individual differences in risk aversion and integrative attitudes.

The first set of hypotheses examined the relationships between individual differences and negotiation strategies. Hypotheses 1a–d examined the incremental variance in negotiation strategy explained by individual differences in risk aversion beyond the effects of contextual power (also controlling for gender and sample). Specifically, Hypothesis 1a proposed more risk averse individuals would be less likely to negotiate. Again, controlling for gender and sample in step 1 of a logistic regression and the respective power variables in step 2, results support Hypothesis 1a. Risk averse individuals were only 0.40 times as likely to enter negotiations (β = −0.92, Wald = 5.85, p < 0.05, see Table 3). Partial support was found for Hypothesis 1b that risk averse individuals who chose to negotiate would be more inclined to adopt accommodating strategies. While an effect was found, it was approaching significance (β = 0.18, t = 1.71, p = 0.09, see Table 4). Risk aversion explains 2.6 per cent incremental variance in use of an accommodating strategy. Hypotheses 1c and 1d, respectively, examined the influence of risk aversion on competing and collaborating strategies. Results indicate that risk aversion explains 3.0 per cent additional variance in use of a competing strategy and 8.4 per cent additional variance in use of a collaborating strategy. Consistent with each of Hypothesis 1c and 1d, risk averse individuals were less likely to compete (β = −0.19, t = 1.97, p < 0.05) or collaborate (β = −0.31, t = 3.42, p < 0.05). Full results are presented in Table 4.

Table 3. Logistic regression regressing decision to negotiate onto risk aversion

	β	SE	Wald	Odds ratio
* p < 0.05 ** p < 0.01.
Step 1
Gender	0.02	0.43	0.00	1.02
Sample	1.78**	0.46	14.74**	5.94
Step 2
Market info.	0.11	0.23	0.42	1.12
Offers	0.71*	0.31	5.13*	2.04
Experience	0.13	0.20	0.42	1.14
Step 3
Risk aversion	−0.92	0.38	5.85*	0.40

Table 4. Hierarchical regressions regressing strategy onto risk aversion

	Competitive	Collaborative	Compromising	Accommodating
* p < 0.05 ** p < 0.01 *** p < 0.10.
Step 1
Gender	−0.03	−0.18***	−0.02	0.04
Sample	−0.21*	−0.05	−0.14	0.05
R^2	0.046***	0.036	0.011	0.004
Step 2
Market info.	0.27**	0.38**	0.28**	−0.12
Offers	−0.07	−0.16***	−0.01	−0.03
Experience	0.29**	0.16***	0.02	−0.25*
ΔR^2	0.157**	0.174**	0.081*	0.08*
Step 3
Risk Aversion	−0.19	−0.31**	−0.17***	0.18***
ΔR^2	0.03*	0.084**	0.028***	0.026***

Hypotheses 2a–b examined the incremental variance in negotiation strategy explained by individual differences in integrative attitudes beyond the effects of contextual power (also controlling for gender and sample). Hypothesis 2a proposed negotiators who hold more integrative perspectives on negotiation will be more likely to negotiate than those with less integrative attitudes. Similar to the analytic strategy outlined in Hypothesis 1a, we control for sample and gender in step 1 and power in step 2 of the logistic regression. Results, presented in Table 5, reveal that consistent with our reasoning, those with more integrative attitudes toward negotiation were in fact 1.82 times more likely to negotiate (β = 0.60, Wald = 4.47, p < 0.05).

Table 5. Logistic regression regressing decision to negotiate onto integrative attitudes

	β	SE	Wald	Odds ratio
* p < 0.05 ** p < 0.01.
Step 1
Gender	0.02	0.43	0.00	1.02
Sample	1.78**	0.46	14.74**	5.94
Step 2
Market info.	0.11	0.23	0.23	1.12
Offers	0.71*	0.31	5.13*	2.04
Experience	0.13	0.20	0.42	1.34
Step 3
Attitudes	0.60**	0.28	4.47**	1.82

Furthermore, Hypothesis 2b maintains that among those who choose to negotiate, those with an integrative perspective will be most likely to adopt a collaborative strategy. To test this hypothesis, we regressed, individually, the remaining four strategies (compete, collaborate, compromise, and accommodate) onto integrative attitudes (controlling for sample and gender in step 1, and power in step 2). Results of this analysis indicate that integrative attitudes only explained significant incremental variance in collaborative strategy (7.3 per cent, see Table 6). Thus, negotiators with integrative attitudes preferred collaborating strategies (β = 0.29, t = 3.15, p < 0.05). This provides full support for Hypotheses 2a and 2b.

Table 6. Hierarchical regressions regressing strategy onto integrative attitudes

	Competitive	Collaborative	Compromising	Accommodating
* p < 0.05 ** p < 0.01 *** p < 0.10.
Step 1
Gender	−0.03	−0.18***	−0.02	0.04
Sample	−0.21*	−0.05	−0.14	0.05
R^2	0.046*	0.036	0.021	0.004
Step 2
Market info.	0.27**	0.38**	0.28**	−0.12
Offers	−0.07	−0.16***	−0.01	−0.03
Experience	0.29	0.16***	0.02	−0.25
ΔR^2	0.157**	0.174**	0.077*	0.08*
Step 3
Attitudes	0.15	0.29**	0.13	−0.09
ΔR^2	0.021	0.073**	0.015	0.007

Hypotheses 3a–b examined the relationships among negotiation strategies and resultant salary increase (controlling for gender and sample effects). Hypothesis 3a predicted the use of a more collaborating strategy would result in increased salary offer. Results provide partial support for this prediction (β = 0.19, t = 1.77, p = 0.08). Negotiators who used more collaborative strategies were able to raise their salary offers; however this effect was only approaching significance (see Table 7). Hypothesis 3b proposed use of a more competing strategy would result in increased salary offer. Results support this proposition (β = 0.34, t = 3.28, p < 0.05, see Table 7).

Table 7. Hierarchical regression regressing salary increase and reactions onto strategies

	Salary increase	ΔR^2	Reactions	ΔR^2
* p < 0.05 ** p < 0.01 *** p < 0.10.
Step 1
Gender	−0.23*		−0.09
Salary	−0.19***		−0.25*
	0.092*		0.069*
Step 2
Competitive	0.34**	0.107**	−0.19*	0.035*
Step 2
Collaborative	0.19***	0.034***	50**	0.217**
Step 2
Compromising	0.17	0.028	−0.43	0.18**
Accommodating	0.03	0.001	0.29**	0.081**
ΔR^2

Hypotheses 4a–c examined the impact of negotiation strategies on reactions to negotiations (controlling for gender and sample effects). Specifically, Hypothesis 4a proposed use of a collaborating strategy would lead to more favorable reactions. Results support this proposition (β = 0.50, t = 5.92, p < 0.05). Hypothesis 4b, that negotiators using a compromising strategy would have more favorable reactions, was similarly supported (β = 0.29, t = 3.10, p < 0.05). Finally, Hypothesis 4c proposed negotiators using an accommodating strategy would have less favorable reactions about negotiations. This hypothesis was also supported (β = −0.43, t = 4.93, p < 0.05). See Table 7 for full results.

To test Hypothesis 5, that negotiation strategy would mediate the relationship between individual differences and negotiation outcomes a series of mediated regression analyses were conducted. The joint significance test, an analytic strategy outlined by MacKinnon, Lockwood, Hoffman, West, and Sheets (2002) was followed for all mediation analyses. The appropriate approach for testing mediation has been a topic of recent debate (MacKinnon et al., 2002; Shrout & Bolger, 2002). In their Monte Carlo study, MacKinnon et al. demonstrate the joint significance test approach provides a better balance of Type I error rates and statistical power compared to more traditional approaches, such as the Baron and Kenny (1986) three-step method. In order for mediation to occur, the joint significance test requires that two conditions be met: (1) The independent variable must predict the mediator and (2) the mediator must predict the dependent variable controlling for the independent variable.

Results of the mediation analyses are presented in Tables 8 and 9. As demonstrated in Tables 8 and 9 and consistent with our I-P-O model, there were two primary input variables examined: Risk aversion (results present in Table 8) and integrative attitudes (results presented in Table 9). Three of the five strategies those were important process variables and examined as mediators of these input variables were: Competitive, collaborative, and accommodating strategies. Compromising was not examined as this strategy was not hypothesized to relate to the individual difference variables examined in the present investigation. Avoid was also not examined as individuals who chose not negotiate did not have the opportunity to raise their salary offer through job negotiations and thus, did not provide us this information. The outputs examined included the two primary dependent variables discussed throughout this paper: Salary increase and reactions to negotiations.

Table 8. Mediation analysis: testing effect of negotiation strategies on attitudes-outcome relationships

	Step 1a	Step 2a	Sobel test
	Collaborative	Reactions
* p < 0.05 ** p < 0.01 *** p < 0.10.
Controls
Sample	−0.18***	−0.09
Gender	−0.05	−0.25*
Independent variable
Integrative Attitudes	0.37**	0.31*
Mediator
Collaborative		0.45**	Z = 3.10**

Table 9. Mediation analysis: testing effect of negotiation strategies on risk-outcome relationships

	Step 1a Competitive	Step 1b Collaborative	Step 1c Accommodate	Step 2a Salary raised	Sobel test	Step 2b Reactions	Sobel test
Note: Standardized regression coefficients (betas) are shown. * p < 0.05 ** p < 0.01 *** p = 0.10.
Controls
Gender	−0.03	−0.18***	0.04	−0.23*		−0.09
Sample	−0.21*	−0.05	0.05	0.19^†		−0.25*
Independent variable
Risk Aversion	−0.31**	−0.41**	0.25**	−0.04		−0.11
Mediators
Competitive				0.36**	Z = 2.36*
Collaborative				0.21***	Z = 1.65***	0.55**	Z = 3.66**
Accommodating						−0.43**	Z = 2.27*

More specifically, in Hypothesis 2b integrative attitudes were found to influence collaborating strategies. We thus examine how collaborating strategies mediate the relationship between integrative attitudes and reactions to negotiations. In Hypotheses 1b–d risk aversion was found to influence competing, collaborating, and accommodating strategies. We in turn examine how competing and collaborating strategies mediate the relationship between risk aversion and salary increase (recall, competing and collaborating explained variance in salary increase, see Hypotheses 3a–b); and also examine how collaborating and accommodating strategies mediate the relationship between risk aversion and reactions to negotiations (recall, collaborating, accommodating and strategies each explained variance in reactions to negotiations, see Hypotheses 4a–b). We examined each path individually. That is, we entered each mediator individually, rather than entering all mediators simultaneously.

Step 1 of mediation is fulfilled as expected given previous hypothesis testing. Controlling for sample and gender, the independent variables each predict their respective mediators. Namely, risk aversion is a significant predictor of each of the aforementioned four strategies (see Table 9 steps 1a–1c), and integrative attitudes are significant predictor of collaborating strategies (see Table 8 step 1a). Next we must determine that the strategies predict salary raised and reactions to negotiation, respectively controlling for risk aversion and integrative attitudes. Tables 8 and 9 provide full results of step 2, as well as supporting Sobel tests, and a summary is provided here. Support was found for our mediation analyses.

As presented in Table 8, use of a collaborating strategy mediated the relationship between integrative attitudes and reactions to negotiations. Follow up Sobel tests support this mediating role (Z = 3.10, p < 0.01). Thus, it appears that negotiators with more integrative attitudes had more favorable reactions to negotiations and this occurred through their use of more collaborative strategies.

As presented in Table 9, the relationship between risk aversion and salary raised was mediated by use of competing strategies (Z = 2.36, p < 0.05) and partially mediated by the use of collaborating strategies. Namely, risk averse individuals were less likely to raise their salaries and this was in part due to their decreased likelihood of adopting competing strategies and to a less extent collaborating strategies (Z = 1.65, p = 0.10). Similarly, the relationship between risk aversion and reactions to negotiation was mediated by collaborating (Z = 3.66, p < 0.01) and accommodating strategies (Z = 2.27, p < 0.05). Risk averse individuals were less likely to have positive reactions toward negotiations and this was in part due to their increased likelihood of adopting an accommodating strategy and decreased likelihood of adopting a collaborative strategy.

Discussion

Consistent with other studies, we found that those who chose to negotiate increased their starting salaries by an average of $5000. Notably, we saw increases in negotiated salaries through the use of competing and collaborating strategies. These approaches are both characterized by an open discussion of one's positions, issues, perspectives, and an exchange of information on how to move forward. Interestingly, beyond their shared variance associated with salary increase, competitive strategies were uniquely associated with salary increases. Competitive approaches to salary negotiations include a strong concern for self-interest and forceful tactics that may include persuasion, threats, and misrepresentation in favor of one's own interests. Compromising and accommodating approaches to negotiation were not associated with negotiated salary gains (or losses).

Use of a collaborative strategy was associated with more positive reactions to the job negotiation process. Those who used accommodating strategies reported more negative reactions to job negotiations. Our study highlights the notion that a win-lose negotiation strategy will earn one the most money in negotiations, but people feel better about their negotiated outcomes when they use “win–win” cooperative approaches that actively take both parties' concerns into consideration.

Individual differences played a significant role in predicting whether or not individuals negotiated, and if so, what strategies they used. Risk averse individuals tended to shy away from negotiating at all. When risk averse individuals did opt to negotiate, they were most inclined to accommodate—a strategy that did not increase their salaries, and left them with feelings of dissatisfaction and unfairness.

Integrative attitudes also impacted choice of negotiation strategy. More integrative-minded individuals engaged in salary negotiation more often than did less integrative-minded individuals, and used a collaborative strategy. In other words, people who believed that negotiation can result in win–win situations relied upon a more active strategies that put the key issues on the table, while still expressed concern for the outcomes of the organization.

One of our most important findings was that negotiation strategies, defined in our study as stylistic approaches to interaction during the salary negotiation process, was the mechanism through which individual difference variables influenced negotiation outcomes. Collaborating strategy mediated the relationship between two individual differences: Risk aversion and integrative attitudes and reactions to negotiations. Those who were risk averse used an accommodating strategy that resulted in more negative reactions to the negotiation process. Furthermore, less risk averse negotiators were more likely to use a competing style that led them to higher salary gains.

Contrary to expectations, a compromising style had less impact on study outcomes and demonstrated lower relationships with individual difference variables. Given that the use of a compromising strategy failed to add significant gains (i.e., financial) to negotiation beyond the use of other strategies (i.e., competing, collaborating) the question of whether compromising represents a distinct factor merits answering. Namely, compromising, like accommodating, represents a strategy in which the negotiator makes large concessions to reach an outcome. Negotiators who adopt accommodating and compromising strategies demonstrate lower levels of concern for their own outcomes according to the Dual Concerns Model (Lewicki et al., 2004) than negotiators who adopt either a collaborating or competing style. An additional CFA was run (in addition to those discussed in the methods section) in which the compromising items were specified to load onto the accommodating factor. Compared to Model 1 (in which compromising items are specified to load onto the compromising factor; see Table 2) the resulting model provided a worse fit to the data (Δχ²(4) = 54.56, p < 0.01). Consistent with models presented in Table 2, any deviation from the measurement model provides a poorer fit to the data. It appears that compromising is a distinct factor. However, exploring more and more varied outcomes may reveal better support for this strategy.

Our research expands the limited field research on job negotiations that has generally been confined to understanding the role of gender in salary negotiations. We found that women were no less likely to negotiate, but those who did were less effective at raising their salaries as compared to men. And though women were more likely to have integrative attitudes about negotiation, they were no more inclined towards a particular negotiation strategy as compared with men. A meta-analysis completed by Walters, Stuhlmacher, and Meyer (1998) that found that women were only slightly more cooperative in negotiations in general. However, Walters et al. (1998) found that when against negotiators using competitive strategies, women were significantly more competitive than men. Thus, it may be that another key factor in selection of negotiation strategy is reaction to a counterpart negotiator's approach to bargaining.

One question raised from these findings is whether or not people using collaborative strategies gained less in monetary outcomes, but were able to realize more gains in non-salary outcomes though bringing more issues into the negotiation. Certainly in academia and often elsewhere, salary is only one of a spectrum of job-related factors that may included in offers and that be negotiable. We collected additional data on non-salary gains in negotiation, including increased non-salary benefits, vacation time, etc. Among people in our study who chose to negotiate, 57.7 per cent received at least one non-salary benefit (e.g., more vacation time, better benefits, etc.). Analyses indicated that people who used an accommodating strategy received less nonmonetary gains than others, while the relationship between nonmonetary gains and collaborative strategies was moderately significant (r = 0.18, p = 0.067). No other strategies had significant relationships with non-salary outcomes.

One explanation is that the primary focus of job negotiations is around raising salaries. Interestingly, there was no relationship between nonmonetary gains and the other outcome variables examined in this study. While the relationship between monetary gains and reactions to negotiations was significant, nonmonetary gains were unrelated to salary increases or to reaction to the negotiation process. This suggests that salary is the salient negotiated outcome for most people during salary negotiations. Based on the argument that starting salary has the longest lasting effects (based on the earlier example, a difference of $600 000+ over the course of a career), people focus on salary because travel, time off, bonuses, etc. do not have the same long-term impact on the individual and these intangibles may be more easily re-visited later.

Job negotiations represent an organization's last opportunity to convince applicants to accept the offer extended to them. Barber (1998) underscores the need for research to examine the effects of post-offer recruitment activities, of which job negotiations would be one. She notes “recruitment does not end when an offer is extended—in fact, it may become especially important at this stage” (p. 149). Thus, more fully understanding the job negotiation process can also provide some needed insight into an important late-state recruitment phenomena that has lasting impact for both the organization and new employee. We only examined the role of the job candidate in salary negotiations, but as a relational process, the organization's negotiating strategies would undoubtedly influence the process and outcomes. Rau and Feinaur (2006) have noted the lack of research on the role of the organizational agents who negotiate salary packages, but have advocated that factors such as organizational agent effort level and perceived ability to influence are important factors in the salary negotiation process.

In addition, our samples focused on the salary negotiation process only for those applicants who accepted job offers. More research on the job negotiation process for the full range of job applicants entering into salary negotiations would be valuable to understand factors that lead to the failure to hire. There are likely situation factors (e.g., better alternative job offers) and organizational factors (poor negotiation process on the behalf of the organization) that may impact failure to hire decisions. Similarly, the literature on job negotiations would be greatly enhanced by investigating both sides of negotiations. Namely, gathering data from both the perspective job seeker and organizational representative would advance understanding of the reciprocal nature of job negotiations.

Limitations

Few field studies of salary negotiations have been undertaken, and while we believe that field studies are the critical next step in testing our experimentally-based negotiations findings, there are limitations associated with this type of retrospective design. We relied on respondents' reflection of their job negotiations, which in some cases happened anywhere from a few months to a few years back. The primary weakness of this approach is the potential impact of memory over time. It is important to note, however, that retrospective designs are common designs used in organizational research to study phenomena such as reactions to performance appraisals, managerial decisions, and recruitment decisions, to name just a few (Aquino, Tripp, & Bies, 2006; Bies & Shapiro, 1987; Bies, Shapiro, & Cummings, 1988; Greenberg, 1986; Holtz & Harold, 2008). In Bies and colleagues' work examining reactions to managerial decisions (Aquino et al., 2006; Bies & Shapiro, 1987) effects were established in a controlled laboratory setting and then replicated in a retrospective field setting, thus providing support for the use of retrospective designs with actual employees. Similarly, retrospective designs are also quite common in the applicant reactions literature (see Bauer et al., 2006 for an example).

Furthermore, surveying negotiators while they were actively involved with the process may raise serious ethical concerns. Namely, asking our respondents questions pertaining to their strategy or with regard to their satisfaction and fairness with the process may effectively interfere with the process, causing them to change their behavior. For instance, a negotiator responding to questions about whether they feel their offer or the process to date is fair, may begin to rethink the terms of the offer given or decide s/he is no longer being treated fairly and reject what was before a satisfactory offer. Similarly, a negotiator presented with items questioning competitive versus accommodating strategies may rethink the tactics s/he is employing and change his/her behavior toward the organization. Thus, the risk that a Hawthorne-type effect might occur by presenting such information in tandem with the actual negotiating process must be weighed along with the limitations of a retrospective design.

Related, the length of time since negotiating the terms of their offer could have an impact on important study mediators and outcomes. We, however, did not collect data that would enable us to control for the time elapsed between negotiation experience and survey completion. This will be an important consideration for future research. Another potential limitation was our sample. We conducted two survey administrations—one among university faculty members and another using part time MBA students who were employed full time by local organizations. The MBA sample contained employees that worked across industries. It is possible that there are different approaches or norms to salary negotiation across industries. We noted that the magnitude of the effects differed very little across samples. Still, because sample differences were found among a few study variables, we controlled for sample for all of our statistical analyses. Furthermore, the similar magnitude of effects across the two samples gives some preliminary evidence for cross-industry generalizability of these findings.

The study methodology used a single-administration, cross-sectional study which may artificially inflate relationships among variables due to common method bias. However, this is a criticism levied against any cross-sectional design. Given the correlations in Table 1 were not overly inflated, and results of Harman's one-factor test indicates that neither a single factor nor a dominant factor accounts for the majority of variance in survey responses, we can be more confident that common method bias is not an underlying cause for study results. Additionally, three of our scales had reliabilities lower than the generally accepted 0.70 level. The compromise and competing strategy scales were among these scales whose coefficient alphas were in the 0.63 and 0.62, respectively. However, a confirmatory factor analysis indicated our strategy scale had fair fit. Furthermore, our scale was an adaptation from the well-validated Rahim (1983) scale. Finally, given the high level of support for the hypotheses involving competition, for instance, would suggest that perhaps our results actually provide a conservative estimate of true effects. Regardless, future research examining job negotiations strategies will be needed to replicate and improve upon these scales.

Yet there were advantages to our sample as well. Respondents were the actual negotiators who had participated in real salary negotiations. Most previous research has relied on student simulations of negotiations. These types of simulated negotiations make it difficult to assess affective-laden outcomes accurately, because of the lack of realistic emotional investment. In sum, we have presented an initial set of evidence that there seems to be meaningful connections between both contextual and individual difference variables as predictors of negotiation strategies and outcomes.

Implications

There are important implications of these findings for both organizations and job seekers. First and most obviously, job seekers must be aware of which strategies and actions will be most effective in terms of salary gains. As we pointed out, just $5000 difference in starting salary may have implications for over $600 000 worth of earnings over one's career.

Our results suggest only active strategies will result in salary gains. Formal programs educating graduating students and individuals who plan to change jobs, on how to engage in such strategies may be especially beneficial. This would serve a dual benefit—forming attitudes about the value creation possibilities available through effective negotiation, and developing a set of skills to enhance actual negotiations.

The results of the present investigation call to light the importance of affective responses to the salary negotiation process. Organizations should be especially cognizant that salary negotiations represent one final opportunity to convince their top candidates to accept their job offer. To the extent the negotiation process is perceived as unfair or the offer as unreasonable, the organization's chances of securing an affirmative response are reduced; decreasing the utility of the entire recruitment process. Moreover, even for those candidates who ultimately accepted the offer despite perceived unfair negotiations, the relationship between employer and employee could be damaged. Future longitudinal research examining the long-term impact of negotiations on commitment, performance, and other related important outcomes will help further elucidate whether the job negotiations has a long lasting impact on the relationship between the employee and organization.

Conclusions

In conclusion, we provided an analysis of how several types of individual differences—personality based risk aversion, attitudes, as well as contextual information influence the use different negotiation strategies in a salary negotiation. We then examined how negotiation strategies influenced salary negotiation outcomes directly, and as mediators of individual differences. This allowed us to extend previous research on the role of individual differences and negotiating strategies in predicting outcomes of a salary negotiation. Our study offers evidence that risk-aversion and integrative attitudes are important predictors of the use of a collaborative approach to negotiation. Collaborative styles, along with competitive strategies, lead to significant gains in salary offers, while collaborative approaches also led to feelings of satisfaction with the negotiated deal and fairness with the negotiated process. Negotiators who accommodated ended up feeling most unsatisfied, and were left with the perception that the process to their negotiated agreements was unfair.

Acknowledgements

We would like to thank Kate Bezrukova, Matt Cronin, and Kevin Rockmann, and Claus Langfred for their comments on an earlier version of this paper.

Biographical Information

Michelle A. Marks is an Associate Professor of management in George Mason University's School of Management. She earned her Ph.D. in Industrial/Organizational Psychology from George Mason University. Dr Marks teaches courses in organizational behavior, leadership, global business and human resource management in executive, MBA, and undergraduate programs. Her research and consulting interests are in the areas of high performance teams, multi-team systems, leadership development, and strategic alliances.

Biographical Information

Crystal Harold received her doctorate in Industrial–Organizational Psychology from George Mason University. She is currently an Assistant Professor in the Department of Human Resource Management in the Fox School of Business and Management. Her research aims to enhance organizational recruitment theory and practice. Specifically, she examines the role of person–environment fit perceptions in the applicant job choice process, the role of employer brand image, and strategies for organizations to improve their recruitment processes. Additionally, she examines issues surrounding applicant and employee reactions to injustice.

Appendix A

Negotiation Strategy

Competition Scale

• (1)
  During negotiation, I tried to persuade the organization to better my offer by threatening to withdraw from the process.
• (2)
  In the negotiation process, I presented information about my past record and qualifications to improve the quality of the offer extended to me.
• (3)
  During negotiations, I made clear the value and benefit I could bring to the organization, in an attempt to influence the process.
• (4)
  While negotiating, I did not take “no” for an answer.
• (5)
  During the negotiation process, if I felt that the organization's offer was unreasonable, I made sure to make my feelings known.
• (6)
  I presented information about the market value of the position for which I was hired.

Collaborate Scale

• (1)
  I tried to negotiate an offer that was acceptable to both me and the organization.
• (2)
  I tried to integrate my interests with those of the organization to come up with an offer supported by both sides.
• (3)
  I tried to work together with the organization to come up with an acceptable offer.
• (4)
  I exchanged accurate information with the organization to come to a joint agreement.
• (5)
  I tried to bring all of our concerns out in the open so that the issues can be resolved in the best possible way.
• (6)
  I collaborated with the organization to come up with an offer acceptable to both of us.
• (7)
  I tried to work with the organization to gain a thorough understanding of their position.

Compromise Scale

• (1)
  I tried to find a middle ground to reach an acceptable offer.
• (2)
  I proposed a middle ground to resolve the differences between our two sides.
• (3)
  I used “give and take” so that compromise could be made.

Accommodate Scale

• (1)
  I initiated job negotiations, but I gave in to the demands of the organization.
• (2)
  To reach an agreement, I allowed more concessions than the organization.
• (3)
  I felt myself trying to accommodate the wishes of the organization.
• (4)
  Though I attempted to negotiate, I found myself going along with much of what the organization initially offered.

Inputs

Integrative Attitudes

• (1)
  Negotiation is almost always a process of one's gain over the other.
• (2)
  A good title to negotiation is a confrontation between sides with opposite goals.
• (3)
  Negotiation is almost always a process in which both sides can benefit simultaneously.

Risk Aversion in Negotiations

• (1)
  I am not willing to take risks when choosing a job or a company to work for.
• (2)
  I prefer a low risk/high security job with a steady salary over a job that offers high risks and high rewards.
• (3)
  I prefer to remain on a job that has problems that I know about rather than take the risks of working at a new job that has unknown problems even if the new job offers greater rewards.
• (4)
  I view risk on a job as a situation to be avoided at all costs.

Market Information

• (1)
  How much did you know about the organization from which you received an offer?
• (2)
  How much did you know about the salary and/or benefits of current employees in the position you applied for?
• (3)
  How much information did you gather about previous offers made by the organization?
• (4)
  How much information did you gather about the market value of your position?

Experience

• (1)
  Prior to this job offer, how many times had you negotiated job offers?
• (2)
  How many full time jobs have you had before taking this position?
• (3)
  How many years of full time work experience have you had before taking this position.