Entitlement - the belief that one deserves a valued resource or positive outcome - is widely regarded as a core feature of narcissism (see, e.g., Akhtar and Thomson 1982; Brown et al. 2009; Dickinson and Pincus 2003; Millon 2011). Accordingly, there is increasing interest in self-report measures of narcissistic entitlement. Some of the most commonly used instruments include the Psychological Entitlement Scale (PES; Campbell et al. 2004) and subscales derived from the Narcissistic Personality Inventory (NPI; Raskin and Terry 1988). However, the existence of different measures of entitlement raises questions about their psychometric strengths and weaknesses as well as their convergent validity. Such information can help researchers make decisions about which instruments should be used in future studies.

The present work further evaluates the PES and the Entitlement/Exploitativeness subscale from the NPI (NPI-EE; Ackerman et al. 2011). Given the potential for ambiguities in the literature, it is important to better understand how these two measures assess narcissistic entitlement (see Miller et al. 2011a for a similar argument). Towards that end, we replicate and extend previous findings concerning these two measures. Moreover, we draw upon the conceptual framework of normal and pathological narcissism forwarded by Pincus and Lukowitsky (2010) to interpret any differences between these measures as this perspective may provide useful insights about narcissistic attributes (see Pincus and Roche 2011). Last, we consider estimates of the internal consistency and test-retest reliability of the PES and NPI-EE to provide additional psychometric information about these instruments.

Conceptualizing Narcissistic Entitlement

When entitlement is considered within the context of narcissistic pathology, we believe that the attribute takes on a distinct set of qualities. Listed as one of the nine diagnostic criteria for Narcissistic Personality Disorder (NPD) in the Diagnostic and Statistical Manual-IV (DSM-IV; American Psychiatric Association 1994), entitlement is defined as “unreasonable expectations of especially favorable treatment or automatic compliance with [an individual’s] expectations” (p. 661). Thus, a critical element of narcissistic entitlement is the idea that the claim to a positive outcome or resource is undeserved or largely unrealistic. We therefore think it is valuable to make a distinction between normal entitlement and narcissistic entitlement (see also Lessard et al. 2011, for a similar argument), a distinction that in many ways parallels the division made by Pincus and Lukowitsky (2010) between normal and pathological narcissism.

In normal entitlement, the needs or demands connected with the claim to a valued resource stem from positive feelings of self-worth and actual accomplishments (Lessard et al. 2011), are at least somewhat realistic (Moses and Moses-Hrushovski 1990), and can potentially be satisfied (Bishop and Lane 2002). The defining feature of this attribute is that there is at least some legitimacy to the claim. Narcissistic entitlement, on the other hand, occurs when the needs or demands associated with the claim are based on unrealistic appraisals (Moses and Moses-Hrushovski 1990), exceed what is due (Grey 1987), and may never be satisfied (Bishop and Lane 2002). In addition, people with a sense of narcissistic entitlement may believe that they deserve special favors that occur explicitly at the expense of others (Lessard et al. 2011). We believe that this second type of entitlement embodies attributes associated with pathological narcissism given its likelihood of producing distress and impairment for others.

One unresolved issue is whether feelings of narcissistic entitlement are more closely connected to expressions of vulnerability, grandiosity, or both. Pincus and Lukowitsky (2010) drew a critical distinction between the grandiose and vulnerable expressions of pathological narcissism. Grandiosity is related to maladaptive self-enhancement strategies (e.g., engaging in grandiose fantasies and exhibitionistic behaviors) whereas vulnerability reflects a fragile self-concept and tendencies toward emotional dysregulation (Pincus and Lukowitsky 2010). Millon (2011) described a variant of entitlement in which the beliefs of special privilege and even a willingness to exploit others ostensibly stem from a self-image of superiority. These entitled individuals are not demanding special treatment from others because they are hurt or vulnerable; rather, they are expecting special treatment because they think they are better than others (thus reflecting a grandiose self-image). This perspective seems to suggest that entitlement has little to do with low self-regard.

On the other hand, certain psychoanalytic and interpersonal theorists have contended that many entitled attitudes are essentially defensive (Bishop and Lane 2002; Coen 1986; Rothstein 1977) as they represent a compensatory reaction to earlier experiences and feelings of deprivation (Bishop and Lane 2002; Coen 1986; Kernberg 2007; Meissner 2008; Moses and Moses-Hrushovski 1990). Negative affectivity and feelings of low self-regard are often tied to the concept of narcissistic entitlement in these writings (Coen 1986; Bishop and Lane 2002; Kris 1990; Meissner 2008; Moses and Moses-Hrushovski 1990; Rothstein 1977). There are also indications that some individuals with NPD have low rather than high self-esteem (Vater et al. 2012) but the empirical connection between narcissism and self-esteem remains controversial (see Bosson and Weaver 2011 and Rosenthal and Hooley 2010 for recent reviews). Nonetheless, there are reasons to suspect that certain expressions of entitlement are linked to vulnerability and low self-regard (see, e.g., Maxwell et al. 2011). Given divergent perspectives about the underlying nature of entitlement, it is important to evaluate whether different measures of entitlement have opposing associations with emotional instability and self-esteem (cf. Dickinson and Pincus 2003).

Measures of Narcissistic Entitlement

As it stands, we suspect that both the grandiose and vulnerable perspectives concerning the expression of entitlement have merit. What is critical is to determine how existing measures of entitlement correspond to these various forms of entitlement described in the literature so as to avoid unnecessary confusion. Although different measures have been developed to assess entitlement, we restrict our primary focus to two measures: the PES (Campbell et al. 2004) and the NPI-EE (Ackerman et al. 2011).

The Psychological Entitlement Scale

We chose the PES because it has become a popular alternative to entitlement scales from the NPI. Indeed, many researchers prefer this measure to the entitlement-based NPI measures in light of its internal consistency and other presumed psychometric strengths (see, e.g., Brown et al. 2009). Campbell et al. (2004) designed the PES to assess, “…a stable and pervasive sense that one deserves more and is entitled to more than others” (p. 31). In support of the criterion-related validity of the PES, Campbell et al. (2004) found that the PES predicted the amount of candy that participants ostensibly took from children, how deserving they were of a salary, their level of greed, their relatively low level of empathy in close relationships, and their observed expression of aggression. Nonetheless, some evidence suggests that the PES may not tap vulnerable expressions of entitlement. For instance, the PES is not linked to higher levels of Neuroticism (Pryor et al. 2008; but see Campbell et al. 2004) and shows a positive but modest relation to explicit self-esteem (Brown et al. 2009; Campbell et al. 2004; Lessard et al. 2011).

The NPI Entitlement/Exploitativeness Scale

Despite the growing popularity of the PES, the NPI is still widely used in the literature (Cain et al. 2008). Indeed, even the developers of the PES continue to use and advocate for the NPI (see Miller et al. 2012) despite criticism of this inventory (Brown et al. 2009; Brown and Tamborski 2011). In light of the widespread popularity of the NPI in social and personality psychology (Cain et al. 2008), we believe it is important to further evaluate the way the NPI captures entitlement. Specifically, we elected to evaluate the newly developed NPI-EE scale because we believe that it represents a relatively “purer” measure of entitlement than other scales derived from the NPI (e.g., the Entitlement scale from Raskin and Terry [1988] or the Exhibitionism/Entitlement scale from Corry et al. [2009]) given that it excludes many items that are not face valid (see Campbell et al. 2004 for this criticism of the Raskin & Terry scale). In addition, much like the PES, the NPI-EE is increasingly used in the literature (e.g., Carpenter 2012; Cater et al. 2011; Hill and Roberts 2012; Malkin et al. 2013; Zeigler-Hill and Wallace 2011), and thus further work on assessing its reliability and validity seems especially important.

When evaluating evidence for the criterion-related validity of the NPI-EE, Ackerman et al. (2011) found that the scale demonstrated consistent links with pathological outcomes despite having a relatively low level of internal consistency. In particular, the NPI-EE is related to maladaptive personality characteristics such as psychopathy (particularly impulsive antisociality), Machiavellianism, emotional instability (i.e., Neuroticism), and antagonism (i.e., low Agreeableness; Ackerman et al. 2011). Moreover, people with higher scores on the NPI-EE report higher levels of anger and aggression (Rosenthal et al. 2011). The NPI-EE is also related to a variety of constructs reflecting psychological distress, such as reduced life satisfaction (Hill and Roberts 2012), diminished psychological health (Rosenthal et al. 2011), and somewhat lower self-esteem (Ackerman et al. 2011; Maxwell et al. 2011; Rosenthal et al. 2011; Zeigler-Hill and Wallace 2011). Such findings are in line with psychoanalytic and interpersonal conceptualizations of narcissistic entitlement as being related to underlying vulnerability and low self-regard.

Present Research

The PES and NPI-EE are increasingly used to assess narcissistic entitlement. Despite certain similarities, previous research shows that the measures diverge in their relations with psychological distress and self-worth. An important aim of the present research is to contextualize these differences within the conceptual framework of normal and pathological narcissism proposed by Pincus and Lukowitsky (2010). In addition, we provide further information regarding each measure’s reliability.

Study 1 evaluates whether the PES and NPI-EE exhibit differential associations with the two newly derived measures of entitlement by Lessard et al. (2011): exploitive entitlement (which we argue captures narcissistic entitlement) and non-exploitive entitlement (which we argue captures more normal or less pathological forms of entitlement). We extend Lessard et al.’s (2011) work by explicitly comparing the PES to the NPI-EE. Study 1 also investigates the Five-Factor Model (FFM) profile of these measures of entitlement to provide a better understanding of their respective nomological networks. Because we have collected these data in the same sample, we are able to quantify the overlap in the FFM profiles for the four entitlement measures via profile or similarity correlations. This provides insight into each measure’s pattern of convergent and discriminant associations with the FFM traits.

Study 2 evaluates the criterion-related validity of the PES and NPI-EE with respect to self-reports of self-esteem and antisocial behaviors. This will further evaluate differences between these measures with respect to self-esteem. As an additional aim, Study 2 replicates the FFM profiles of the PES and NPI-EE to determine which patterns of association are robust. This kind of replication is important because it bolsters claims regarding the differential nomological networks for the PES and NPI-EE.

Last, in Study 3 we investigate the test-retest reliability or dependability (see Watson 2004) of these measures. McCrae et al. (2011) demonstrated that test-retest reliability/dependability is a better predictor of validity than internal consistency. This is an important observation given that many of the criticisms of the NPI-based entitlement measures (Campbell et al. 2004; Lessard et al. 2011) focus on their comparatively low alpha coefficients (e.g., α for NPI Entitlement scale = .46 to .48 in del Rosario and White 2005). This final study therefore provides further important information when attempting to evaluate the two measures of entitlement.

Study 1

Study 1 evaluated the nomological networks for the PES and NPI-EE. As was argued earlier, we suspect that the two scales are assessing somewhat different forms of entitlement. In particular, the NPI-EE may more effectively tap into the vulnerable expression of narcissistic entitlement than the PES. To evaluate associations with different forms of entitlement, we investigated how the PES and NPI-EE are associated with two recently developed measures of entitlement by Lessard et al. (2011): Exploitive Entitlement and Non-Exploitive Entitlement. Whereas the Exploitive Entitlement scale captures beliefs of special privilege that impinge upon the rights of others, the Non-Exploitive Entitlement scale captures beliefs of deservedness that do not infringe others’ rights. Despite their divergent nomological networks, Lessard et al. (2011) found that both forms of entitlement correlate positively with the PES (r = .51 for Exploitive Entitlement; r = .43 for Non-Exploitive Entitlement) and the NPI Entitlement scale (r = .34 for both Exploitive and Non-Exploitive Entitlement). It appears that the PES and NPI Entitlement scale are tapping into both variants of entitlement. Although we expect to obtain similar results for the PES, we are less certain about what to expect for the NPI-EE given that it deviates somewhat from the NPI Entitlement scale.

We expect that the PES, NPI-EE scale, and Non-Exploitive Entitlement scale from Lessard et al. (2011) should exhibit a differential pattern of relations with the attributes captured by the Five Factor Model. The FFM trait that is most commonly associated with narcissism (and likely relevant for entitlement) is low Agreeableness (Corbitt 2002; Trull and McCrae 2002). Further, Samuel and Widiger (2008) reported that the largest links between measures of NPD and the FFM occurred for the domain of Agreeableness (weighted effect size = −.37). We expect that the PES and NPI-EE should both be negatively correlated with Agreeableness. We do not expect a strong relation between Non-Exploitive Entitlement and Agreeableness.

With respect to Neuroticism, the psychoanalytic and interpersonal perspectives we outlined earlier suggest that individuals with higher levels of entitlement should exhibit personality features related to Neuroticism (e.g., anger, vulnerability). However, this claim is controversial. Indeed, Samuel and Widiger (2008) did not report strong links between Neuroticism and NPD (weighted effect sizes of .10). An important concern, however, is that measures of NPD tend to overemphasize the grandiose features of narcissistic dysfunction discussed in the clinical literature at the expense of the vulnerable features (Cain et al. 2008). Thus, there is value in evaluating this issue with respect to specific measures of entitlement. We expect that the NPI-EE will be positively related to Neuroticism; however, we do not expect to find a relation between Neuroticism and the PES.

Finally, links with the other FFM traits are not predicted to be as strong for the PES or NPI-EE. Corbitt (2002) noted that “the Extraversion domain is not theoretically central to narcissism” (p. 297) and neither Extraversion nor Openness were strongly related to NPD in the Samuel and Widiger (2008) meta-analysis (weighted effect sizes of .09 and .07, respectively). On the other hand, various surveys of clinicians or personality disorder experts have shown that certain facets of Extraversion (e.g., assertiveness; Lynam and Widiger 2001; Samuel and Widiger 2004; Thomas et al. 2012) and Openness (e.g., actions; Lynam and Widiger 2001) are believed to be associated with NPD. The link between Conscientiousness and narcissism is also somewhat unclear. Although there are arguments that impulse control deficiencies are central to narcissistic pathology (Vazire and Funder 2006), Samuel and Widiger reported a meta-analytic estimate of the correlation between NPD and Conscientiousness that was fairly small (weighted effect size of −.10).

Method

Participants and Procedure

Participants included 436 undergraduate college students (51.8 % female) that completed the study in partial fulfillment of course or extra credit in psychological courses. The majority of the participants were White (86.5 %; three participants did not report their ethnicity) and were 18 years old (25.7 %), 19 years old (28.4 %), 20 years old (21.3 %), or 21 years old (14.7 %). Participants completed measures of entitlement and the FFM traits online.

Measures

NPI Entitlement/Exploitativeness

We used the 40-item Narcissistic Personality Inventory (NPI; Raskin and Terry 1988) to compute the NPI-EE (4 items). Example items included, “I will never be satisfied until I get all that I deserve” (narcissistic response) versus “I take my satisfactions as they come” (non-narcissistic response); and “I insist upon getting the respect that is due me” (narcissistic response) versus “I usually get the respect that I deserve” (non-narcissistic response). Each participant’s score was the average of her or his responses, with higher scores reflecting higher levels of the corresponding trait (M = 0.22, SD = 0.25; α = .46, average inter-item r = .18).

Psychological Entitlement

Psychological entitlement was assessed using the measure created by Campbell et al. (2004). Participants responded to nine statements using a scale that ranged from 1 (Strong disagreement) to 7 (Strong agreement). Example items included, “Great things should come to me” and “I demand the best because I’m worth it.” Each participant’s score was the mean of these items, with higher scores reflecting higher levels of entitlement (M = 3.38, SD = 1.15; α = .88, average inter-item r = .46).

Exploitive and Non-Exploitive Entitlement

Developed by Lessard et al. (2011), this questionnaire is comprised of two subscales that assess separate variants of entitlement. The five-item Non-Exploitive Entitlement subscale assesses beliefs of deservedness that do not impinge upon the rights of others (M = 3.32, SD = 0.66; α = .74, average inter-item r = .36). Example items include, “I deserve to be treated with respect by everyone” and “I am entitled to get into the career that I want.” The seven-item Exploitive Entitlement subscale assesses beliefs of special privilege that are possessed at the expense of others (M = 2.30, SD = 0.62; α = .82, average inter-item r = .39). Example items include, “I shouldn’t have to work as hard as others to get what I deserve” and “Because of the things I have been through personally, others should cut me a break in life.” Participants responded to each statement using a 5-point Likert-type scale that ranged from 1 (Strongly disagree) to 5 (Strongly agree). Participants’ scores for each subscale were the average of the items, with higher scores indicating higher levels of the corresponding construct.

The FFM Traits and Facets

Participants completed a short measure of the 30 FFM facets using 120 items from the International Personality Item Pool (IPIP; Johnson 2011).Footnote 1 Participants responded to each statement using a 5-point Likert-type scale that ranged from 1 (Strongly disagree) to 5 (Strongly agree). There were six subscales for each broader trait to capture the corresponding facets; all subscales consisted of four items each. The facet subscales for each domain were averaged to produce a measure of the broader domain. The trait domain of Agreeableness (M = 3.60, SD = 0.41; α = .68) was comprised of the following facets: trust (α = .86), morality (α = .75), altruism (α = .73), cooperation (α = .68), modesty (α = .70), and sympathy (α = .67). The trait domain of Conscientiousness (M = 3.65, SD = 0.48; α = .77) was comprised of the following facets: self-efficacy (α = .75), orderliness (α = .84), dutifulness (α = .66), achievement striving (α = .78), self-discipline (α = .70), and cautiousness (α = .90). The trait domain of Extraversion (M = 3.42, SD = 0.49; α = .76) was comprised of the following facets: friendliness (α = .79), gregariousness (α = .78), assertiveness (α = .87), activity level (α = .74), excitement (α = .73), and cheerfulness (α = .79). The trait domain of Neuroticism (M = 2.82, SD = 0.55; α = .80) was comprised of the following facets: anxiety (α = .81), anger (α = .85), depression (α = .86), self-consciousness (α = .72), immoderation (α = .71), and vulnerability (α = .78). Finally, the trait domain of Openness (M = 3.29, SD = 0.43; α = .65) was comprised of the following facets: imagination (α = .74), artistic interests (α = .78), emotionality (α = .62), adventurousness (α = .64), intellect (α = .74), and liberalism (α = .65).

Results and Discussion

Table 1 presents correlations between the entitlement measures. To begin, we more or less replicated Lessard et al.’s (2011) findings by demonstrating that the PES was linked to both the Non-Exploitive and Exploitive Entitlement scales and that the magnitudes of these correlations did not significantly differ.Footnote 2 Table 1 also shows that the NPI-EE demonstrated significant links with both of Lessard et al.’s (2011) entitlement scales. Of note, the relation between NPI-EE and Exploitive Entitlement was significantly stronger than the relation between NPI-EE and Non-Exploitive Entitlement, t = 2.34, p = .020. Nevertheless, the magnitude of the relations between Lessard et al.’s (2011) entitlement scales and the NPI-EE were considerably smaller than those found with the PES.

Table 1 Zero-order correlations between entitlement measures in Study 1
Full size table

Table 2 presents correlations between the entitlement measures and the FFM domains and facets. Given the large number of correlations being assessed, we place our focus on effect sizes as opposed to statistical significance. We therefore limit our attention to those correlations ≥ |.20|. Figure 1 presents the measures’ FFM profiles on the same graph to facilitate comparisons. The solid black horizontal bars depict the correlation threshold of |.20|.

Table 2 Five-Factor Model (FFM) correlation profiles for the entitlement measures in Study 1(and Study 2)
Full size table
Fig. 1
figure 1

Five-Factor Model Profiles for Entitlement Measures in Study 1. Note. PES = Psychological Entitlement Scale. Lessard Non-Exp = Lessard et al. (2011) Non-Exploitive Entitlement Scale. Lessard Exp = Lessard et al. (2011) Exploitive Entitlement Scale. NPI-EE = Entitlement/Exploitativeness scale from the Narcissistic Personality Inventory

Full size image

Consistent with the contention that antagonism is a central personality element of narcissistic entitlement, the PES, NPI-EE, and Exploitive Entitlement scale had medium to large relations with Agreeableness (i.e., r’s = −.37 to −.52; see Cohen 1988). In contrast, the Non-Exploitive Entitlement scale was weakly related to Agreeableness, consistent with the claim that it assesses a less noxious form of entitlement (see Lessard et al. 2011). At the facet level, participants with higher scores on the PES reported possessing lower levels of morality, cooperation, and modesty. Of note, the negative correlation between the PES and modesty was especially pronounced. Participants with higher scores on the NPI-EE similarly reported lower levels of these same facets, along with lower levels of trust, altruism, and sympathy.

As Table 2 shows, neither the PES nor the Non-Exploitive Entitlement scale was substantially related to Neuroticism. In contrast, the NPI-EE and Exploitive Entitlement scales showed small to medium positive relations with Neuroticism. Examination of the facets shows that these connections were primarily driven by the positive links with the anger facet (although the correlations among NPI-EE, depression, and vulnerability were elevated, they did not surpass the |.20| threshold). These results suggest that the NPI-EE (and perhaps also the Exploitive Entitlement scale from Lessard et al. 2011) are capturing a form of entitlement that has some connection with emotional instability and perhaps vulnerability.

As expected, the measures believed to tap into narcissistic entitlement (i.e., the PES, NPI-EE, and Exploitive Entitlement) were not substantially related to the FFM traits of Extraversion and Openness or their facets. Interestingly, the Non-Exploitive Entitlement scale exhibited a moderately sized positive correlation with Extraversion. Examination of the facets revealed that participants with higher levels of Non-Exploitive Entitlement reported having greater levels of friendliness and excitement. Thus, this trait was linked to more positive affect and even a pleasant engagement with other people—qualities not typically associated with socially noxious personality attributes.

Finally, there were no substantial connections between the entitlement measures and the broad trait of Conscientiousness. It should be noted, however, that there were a few connections with the facets of Conscientiousness that crossed the |.20| threshold. Specifically, the NPI-EE and Exploitive Entitlement scale were both negatively correlated with the dutifulness facet. Further, the Exploitive Entitlement scale was connected with lower levels of cautiousness.

As an additional way to understand the differences between the FFM profiles for the PES and NPI-EE, we performed tests of the differences in the magnitudes of the correlation coefficients of these scales with the FFM traits and facets. Because there were a large number of tests, we focus on those differences significant at p < .01 in the text (differences significant at p < .05 are also noted in Table 2). As Table 2 shows, at the broader domain level, the NPI-EE exhibited a stronger negative connection with Agreeableness and a stronger positive connection with Neuroticism than the PES. In terms of facets, the NPI-EE had stronger connections with Trust, Altruism, Cooperation, Dutifulness, Friendliness, Cheerfulness, and Depression than the PES. In contrast, the PES had stronger connections with Modesty and Intellect than the NPI-EE.

To more precisely quantify the similarity of the FFM profiles for each measure of entitlement, we computed correlations between the profiles using what Samuel and Widiger 2008) refer to as “similarity correlations” (see p. 1331). Table 3 presents this information using both Pearson correlations and intraclass correlations (ICCs) above and below the diagonal, respectively. Because we were interested in the degree of similarity in the magnitude of the correlations comprising the FFM profiles in addition to the consistency of the shape of the FFM profiles, we emphasize the ICCs as they combine information about level similarity (i.e., the average magnitude of the zero-order correlation with the FFM facet in the current example) and shape similarity (i.e., the rank-order consistency of the pattern of zero-order correlations with the FFM facets).

Table 3 Similarity Correlations between the Five-Factor Model (FFM) correlation profiles for Entitlement Measures
Full size table

As can be seen in the ICCs in Table 3, the FFM profiles for the entitlement measures generally converged. The one exception to this trend was the FFM profile for the Non-Exploitive Entitlement scale. Indeed, the ICCs between the FFM profile for this scale and the other entitlement measures were relatively low. It is therefore noteworthy that the ICC between the Non-Exploitive Entitlement scale and the PES was comparably high. In other words, the configuration of basic personality traits found for the PES aligns to some extent with the configuration found for a measure that assesses a seemingly adaptive variant of entitlement (Lessard et al. 2011). On the other hand, the FFM profile for the PES was also aligned with the FFM profiles for the Non-Exploitative Entitlement and NPI-EE scales, suggesting a pattern of correlates that also tracks with a relatively more maladaptive variant of entitlement. Finally, the ICC between the FFM profiles for the NPI-EE and the Exploitive Entitlement scale was especially strong, suggesting a reasonably close level of correspondence between their respective patterns of association with general personality traits.

Study 2

Study 2 replicates the FFM profiles of the PES and NPI-EE found in Study 1. Study 2 also provides further evidence for the criterion-related validity of the PES and NPI-EE. A critical question concerning the utility of any measure of narcissistic entitlement is whether it is able to predict theoretically relevant criteria. Accordingly, we investigated links between these two entitlement measures and several outcomes relevant to the construct of narcissistic entitlement: antisocial behavior, empathy, cheating, and self-esteem.

Brown et al. (2009) argued that empathy deficits and other interpersonal characteristics connected with narcissism (e.g., exploitation) likely result from entitled beliefs. Indeed, the entitled individual may be so consumed with his or her own needs and “rightful” compensation that it becomes difficult to understand how this may conflict with the needs of others (Bishop and Lane 2002). Recent research finding a link between entitlement and selfish behavior appears to support this view (Zitek et al. 2010). We expect that both forms of entitled beliefs (i.e., grandiose and vulnerable) should be related to lower levels of empathy and higher levels of antisocial behaviors.

An excessive sense of entitlement can also be enacted in the societal, as opposed to the interpersonal realm (Moses and Moses-Hrushovski 1990). According to Brown et al. (2009), “…entitlement appears to reflect an overt rejection of social norms, which suggests that people high in entitlement are likely to ignore moral or ethical prohibitions against deliberative cheating” (p. 960). In other words, rules do not apply to individuals with a sense of narcissistic entitlement as they believe they deserve exceptions from the conventional rules that govern behavior. We therefore expect both forms of entitlement to be related to greater cheating.

Several authors have suggested that feelings of entitlement can serve as a defense against internal conflicts involving shame, guilt, rage, and depression (Moses and Moses-Hrushovski 1990; Rothstein 1977). In addition to being linked to feelings of anger and destructiveness (Coen 1986; Bishop and Lane 2002), attitudes of entitlement are also linked to feelings of worthlessness and defectiveness in those clinical writings that emphasize vulnerability (Coen 1986; Moses and Moses-Hrushovski 1990; Rothstein 1977). In line with this, Dickinson and Pincus (2003) argued that “…grandiose narcissistic individuals are less susceptible than their vulnerable peers to the chronic emotional consequences of threats to entitled expectations (e.g., distress, lowered self-esteem, interpersonal fearfulness)” (pp. 200-201). We therefore propose that a measure capturing a somewhat more vulnerable expression of entitlement (i.e., the NPI-EE) will be negatively associated with self-esteem.

Method

Participants and Procedure

The sample consisted of 497 undergraduate college students (51.3 % female). The majority of the participants were White (84.9 %; one participant did not report her or his ethnicity) and were 18 years old (26.4 %), 19 years old (25.8 %), 20 years old (20.7 %), or 21 years old (18.5 %; one participant did not report her or his age). All participants completed questionnaires online pertaining to narcissism, the FFM traits, self-esteem, antisocial behavior, empathy, and cheating in partial fulfillment of course or extra credit for their psychology courses. Data from this sample are also reported in Burt et al. (2012); however, the goals of this study are distinct from the Burt et al. (2012) report which focused on the correlates of social aggression. Moreover, the Ackerman et al. (2011) scale was not reported in the Burt et al. (2012) investigation.

Measures

Entitlement

Participants completed the PES developed by Campbell et al. (2004) using the same scale format and scoring method described in Study 1 (M = 3.49, SD = 1.16; α = .89, average inter-item r = .47). Likewise, participants completed the NPI-EE (M = 0.20, SD = 0.25; α = .52, average inter-item r = .21).

The FFM Domains and Facets

The 120-item International Personality Item Pool (IPIP; Johnson 2011) was again administered to participants using the same scale format and scoring method described in Study 1 (Agreeableness: M = 3.56, SD = 0.44, α = .86; Conscientiousness: M = 3.61, SD = 0.47, α = .89; Extraversion: M = 3.43, SD = 0.47, α = .88; Neuroticism: M = 2.79, SD = 0.51, α = .88; Openness: M = 3.26, SD = 0.43, α = .81). Alphas for the subscales ranged from .56 (for Adventurous) to .86 (for Cautiousness).

Self-Esteem

The Rosenberg (1965) self-esteem scale was used to assess participants’ trait levels of explicit self-esteem. Participants responded to each of the 10 items using a 5-point scale that ranged from 1 (Strongly Disagree) to 5 (Strongly Agree). Each participant’s score was the average of these items, with higher scores indicating higher levels of self-esteem (M = 3.72, SD = 0.64; α = .89).

Antisocial Behavior

We used the Subtypes of Antisocial Behavior (STAB) Questionnaire (Burt and Donnellan 2009) to assess antisocial behavior. Participants were asked to respond to each statement on a 5-point scale that ranged from 1 (Never) to 5 (Nearly all the time). Subscales within the measure assess rule-breaking behavior (e.g., “Stole property from school or work”), social aggression (e.g., “Made negative comments about other’s appearance”), and physical aggression (e.g., “Threatened others”). Because the three subscales yielded relatively similar results in the current study,Footnote 3 we decided to use a total score that represented a broad-band measure of antisocial behavior. Each participant’s score was the average of the items, with higher scores reflecting higher levels of self-reported antisocial behavior (M = 2.13, SD = 0.60; α = .93).

Empathy

Empathy was assessed using the Empathy topical module from the General Social Survey (Smith et al. 2011). Participants responded to seven statements on a 5-point scale that ranged from 1 (Does not describe me very well) to 5 (Describes me very well). Example items include, “I often have tender, concerned feelings for people less fortunate than me” and “When I see someone being treated unfairly, I sometimes don’t feel very much pity for them” (reverse coded). Each participant’s score was the average of these items, with higher scores indicating higher levels of empathy (M = 3.76, SD = 0.67; α = .82).

Cheating

Developed for the current study, the cheating scale consisted of three face valid items that assessed academic cheating. Items included, “I have cheated on school tests”, “I have handed in a school essay that I copied from someone else in the class”, and “I have plagiarized material for a school assignment using the Internet.” Participants responded to each statement on a 5-point scale that ranged from 1 (Strongly disagree) to 5 (Strongly agree). Each participant’s score was the mean of the three items, with higher scores indicating more cheating (M = 1.93, SD = 0.85; α = .70).

Results and Discussion

Table 2 shows the FFM profiles for the PES and NPI-EE based on the data from Study 2. The general pattern of correlations for each measure replicated the results of Study 1. In particular, the connections between the PES and Agreeableness (especially the facets of morality, cooperation, and modesty) were duplicated. Similarly, the connections between the NPI-EE and Agreeableness (and all of its facets), the Dutifulness facet from Conscientiousness, and Neuroticism (the anger facet especially) were repeated. Like Study 1, we performed tests of the differences in the magnitudes of the correlation coefficients of these scales with the FFM domains and facets (differences significant at p < .05 and p < .01 are noted in Table 2). As can be seen, the differences between the PES and NPI-EE in terms of trust, altruism, modesty, Conscientiousness, dutifulness, Extraversion, friendliness, cheerfulness, Neuroticism, anger, and depression were duplicated.

Table 4 presents the zero-order correlations between the study variables. The NPI-EE was again moderately correlated with the PES. Table 4 also shows both the PES and NPI-EE were positively related to antisocial behavior and cheating and negatively related to empathy. Tests of dependent correlations revealed that the NPI-EE’s association with empathy was stronger than the relation observed for the PES (t = −2.85, p = .005), whereas the two measures’ relations with the STAB total score and cheating were not significantly different from each other (STAB total score: t = 1.70, p = .089; cheating: t = 1.65, p = .10). Further, the NPI-EE was negatively related to self-esteem whereas the PES was positively related to self-esteem. A test of dependent correlations confirmed that these two coefficients were significantly different from each other, t = −5.09, p < .001.

Table 4 Zero-order correlations between variables in Study 2
Full size table

To obtain a more precise understanding of the unique relations between these two entitlement measures and the criterion variables, a series of multiple regression analyses were performed in which each of the criteria was regressed upon the PES and NPI-EE (see Table 5). The analyses show that both measures of entitlement statistically predict greater levels of antisocial behavior even when controlling for their overlap. In contrast, lower levels of empathy and greater levels of cheating are uniquely predicted by the NPI-EE but not the PES.

Table 5 Multiple regression analyses with entitlement measures predicting theoretically relevant criteria
Full size table

Study 3

Study 3 evaluated the dependability of the PES and NPI-EE. Short-term test-retest reliability or dependability captures the extent to which scores from the same scale assessed across two time points are similar to one another. Watson (2004) suggested that a 2-week interval balances the assumption of trivial true change in the underlying construct with the need to allow memory effects to diminish. We include the Rosenberg (1965) self-esteem scale as a point of comparison. This information about dependability is important in light of the previous evidence for the criterion-related validity of the two measures of entitlement. Indeed, the results for the NPI-EE might be surprising in light of its relatively low alpha coefficients in previous studies. Recall that internal consistency essentially captures the degree to which items within a scale cohere (with the caveat that alpha coefficients tend to increase with scale length).

Past work has shown that the PES demonstrates satisfactory test-retest reliability (r = .72 for 1 month and r = .70 for 2 months; Campbell et al. 2004). Although there is currently no information about the dependability of the NPI-EE, research has shown that the test-retest reliability for the NPI Entitlement scale was .57 for 13 weeks (del Rosario and White 2005). One caveat in drawing comparisons is that these two measures differ in terms of their internal consistency and low alpha coefficients can attenuate retest coefficients. Accordingly, Study 3 investigates short-term retest reliability coefficients for the PES and NPI-EE across the same 2-week interval with explicit attention to their levels of internal consistency. In line with previous research (Campbell et al. 2004), we expect that the PES will demonstrate strong dependability. Further, given evidence of the criterion-related validity of the NPI-EE, we expect that the test-retest reliability coefficient for the NPI-EE will be larger than its index of internal consistency.

Method

Participants and Procedure

Two-hundred and eleven undergraduate college students completed the first wave of data collection. Of these participants, 150 completed the follow-up approximately 2 weeks later. Unfortunately, seven participants did not provide the same email addresses at both time points so their data could not be linked. In addition, one participant reported that he/she did not answer the questions honestly, and so his/her responses were removed. This left a complete sample of 142 undergraduate college students (75.4 % female), the majority of which were 19 years old (29.6 %), 20 years old (26.8 %), or 21 years old (23.2 %); the sample was predominantly White (76.1 %). Participants completed the PES, NPI-EE, and self-esteem scale at both waves.

We compared participants who completed both waves with participants who only completed the first wave on demographics and the main study variables. Analyses revealed that participants did not differ on ageFootnote 4 (Fisher’s exact test = 10.16, p = .065),Footnote 5 gender (χ 2 (1) = 1.49, p = .223), the PES (only completed first wave: M = 3.96, SD = 0.92; completed both waves: M = 3.72, SD = 0.99; t (206) = 1.66, p = .10), the NPI-EE (only completed first wave: M = 0.25, SD = 0.28; completed both waves: M = 0.23, SD = 0.25; t (208) = 0.51, p = .614), or self-esteem (only completed first wave: M = 3.83, SD = 0.66; completed both waves: M = 4.00, SD = 0.62; t (208) = −1.84, p = .067). However, participants differed on ethnicity, Fisher’s exact test = 15.55, p = .007. Z-tests (with Bonferroni adjusted p-values) comparing the column proportions showed that a significantly higher proportion of participants who self-identified as White Americans were present in the sample subset that completed both waves (p = .76) relative to the subset that only completed the first wave (p = .58). Similarly, a significantly lower proportion of participants who self-identified as Multi-Racial/Other were present in the subset that completed both waves (p = .04) relative to the subset that only completed the first wave (p = .12).

Results and Discussion

Table 6 presents descriptive statistics for each of the scales at each wave. The internal consistency coefficients for all of the scales except the NPI-EE were adequate using the conventional rule of thumb (i.e., an α ≥ .70). Moreover, the test-rest reliability coefficients for the PES (r = .83, p < .01) and Rosenberg self-esteem scale (r = .80, p < .01) were large in size and quite comparable in terms of their magnitudes. The size of the test-retest reliability coefficient for the NPI-EE (r = .59, p < . 01) was lower; tests of the differences between independent correlations confirmed that the dependability coefficient for the NPI-EE was significantly smaller than the dependability coefficients for the PES and Rosenberg self-esteem scale (ps < .05). Although the retest reliability coefficient for the NPI-EE was not as large as the other scales, it is important to note that the coefficient itself was larger than its own index of internal consistency at both waves. Nonetheless, when internal consistency and retest reliability are considered together, it is clear that the PES functions better than the NPI-EE.

Table 6 Descriptive statistics for time 1 and time 2 Study 3 measures
Full size table

General Discussion

Entitlement is considered to be one of the core features of narcissistic pathology (e.g., Brown et al. 2009; Pincus and Lukowitsky 2010; Trull and McCrae 2002). Therefore, researchers who study narcissism need psychometrically adequate measures of narcissistic entitlement. One complicating factor is that different existing measures of entitlement may tap different aspects of the construct. In line with this suggestion, the present research suggests that two entitlement measures—the PES and the NPI-EE—capture somewhat different variants of entitlement. Whereas the PES ostensibly captures a predominantly grandiose expression of narcissistic entitlement, the NPI-EE seems to assess a somewhat more vulnerable expression. Below we elaborate on some of the major findings and discuss some implications for the measurement of entitlement.

Are the Two Entitlement Measures Assessing the Same Construct?

The present research provides further evidence that the PES and NPI-EE do not tap the construct of entitlement in the same way. At a most basic level, if the two measures were perfectly exchangeable, we would expect: 1) strong convergent correlations; and 2) similar patterns of convergent/discriminant validity when evaluating criterion-related correlations. In practice, the correlation between the measures wavered around .40 across Studies 1 and 2 (if the correlations are disattenuated for measurement error, the correlation between the two measures becomes .58 in Study 1 and .65 in Study 2). These findings are generally consistent with the correlations found for the PES and NPI-EE in previous research (e.g., r = .36 in Ackerman et al. [2011]; r = .48 for White participants in Zeigler-Hill & Wallace [2011]). These correlations suggest that the measures may be assessing somewhat different elements of entitlement. This insight is also consistent with analyses focused on the respective nomological networks for these two measures.

The PES seems to demonstrate a pattern of correlates that is aligned with what we consider to be the more grandiose elements of narcissism. For example, the PES correlated strongly with the Exploitive Entitlement subscale by Lessard et al. (2011) and was also characterized by higher levels of antagonism (immodesty in particular) and antisocial behaviors. The similarity correlation between the FFM profiles for the PES and Exploitive Entitlement scale was substantial (Pearson r = .78; ICC = .75) indicating some correspondence in the two scales’ relations with general personality traits. Evidence also emerged, however, to suggest that the PES captures elements of a construct that we have termed normal or non-pathological entitlement. Indeed, the correlation between Non-Exploitive Entitlement and the PES was considerable (r = .58). The similarity correlation between the FFM profiles for PES and Non-Exploitive Entitlement was also relatively substantial (Pearson r = .82; ICC = .77). The PES also had a modest positive association with self-esteem. Taken together, these findings suggest that the PES assesses a mixture of somewhat normal and narcissistic attitudes linked with entitlement (see also Lessard et al. 2011).

In contrast to the PES, we predicted that the NPI-EE would show a pattern of correlates that suggested some underlying vulnerability and emotional instability. Results were somewhat (but not completely) consistent with this proposal. The FFM profile for the NPI-EE was characterized by lower Agreeableness (especially low morality and cooperation), lower dutifulness (a facet of Conscientiousness), and higher Neuroticism (anger in particular). The NPI-EE was also linked to somewhat lower levels of self-esteem and lower levels of empathy along with higher levels of antisocial behavior and cheating. We found that although the NPI-EE was significantly related to both normal and narcissistic entitlement, its relation with narcissistic entitlement was stronger. Further, the NPI-EE and Lessard et al. (2011) narcissistic entitlement scale had similar patterns of association with the FFM attributes. Indeed, compared to the PES, the similarity correlations between the FFM profiles for the NPI-EE and the two Lessard et al. (2011) scales were notably different. This result suggests to us that the NPI-EE is a purer index of narcissistic entitlement rather than a mixture of normal and narcissistic entitlement. We should note, however, that the NPI-EE’s associations with self-esteem as well as Neuroticism and its facets were perhaps not as large as one would have expected for a form of entitlement that is supposedly driven by vulnerability. Thus, we can only conclude that the NPI-EE captures more vulnerability than the PES.

Recommendations for Assessing Narcissistic Entitlement

Recent psychometric discussions have pointed to the utility of discussing reliability in more fine-grained terms, especially when considering internal consistency and test-retest reliability or dependability (cf. McCrae et al. 2011). One of the strengths of Study 3 was that it used the same time interval of 2 weeks to assess the dependability of the PES and NPI-EE. Consistent with previous research Campbell et al. (2004), Study 3 showed that the test-retest reliability coefficient of the PES was substantial. The test-retest reliability coefficient of the NPI-EE was not as high as the PES suggesting that the PES is more dependable than the NPI-EE. However, the dependability of the NPI-EE scale was still an improvement over its own index of internal consistency. This may help to explain why the NPI-EE tends to perform better than expected when predicting relevant criteria (see McCrae et al. 2011).Footnote 6

Past work generally finds that the PES shows greater internal consistency than the NPI-EE (e.g., Ackerman et al. 2011; Zeigler-Hill and Wallace 2011). The current research replicates this pattern of findings. A critical issue is that the NPI-EE only consists of four items that use a forced-choice response format. Given that coefficient alpha is a function of the average inter-item correlation coefficient and the number of items, these factors may be responsible for the consistently low alpha coefficients reported across studies. It is plausible that switching to a Likert-type scale response format and increasing the number of items would help to improve its internal consistency. However, these changes might alter other psychometric properties of the measure, and so we suggest that any researchers who alter the scale should also investigate its corresponding changes in dependability and criterion-related validity.

Our current view informed by this study and previous investigations (Lessard et al. 2011; Pryor et al. 2008) suggests that the PES reliably assesses a mixture of normal entitlement and narcissistic entitlement emphasizing grandiosity. It may be that certain items on the scale (e.g., “Great things should come to me”) capture both normal and grandiose entitlementFootnote 7 whereas other items (e.g., “I feel entitled to more of everything”) reflect more uniquely grandiose expressions of narcissism.Footnote 8 In contrast, when compared to the PES, the NPI-EE seems to assess a form of entitlement that has some connections with vulnerability.Footnote 9 The major downside to this scale is that it has low internal consistency as has now been reported in several independent investigations. Thus, although researchers may wish to use the PES instead of the NPI-EE (or vice versa), we believe researchers should make this decision with care. Given that both scales are relatively brief, it is would seem possible to administer them both to provide more complete coverage of the entitlement construct until improved scales are developed.

In light of the theoretical distinctions between non-exploitative entitlement and exploitative entitlement proposed by Lessard et al. (2011) and the current focus on narcissistic and normal entitlement, we suggest there is a need to further refine existing measures of entitlement. Most notably, future research would benefit from expanding the PES item pool and subsequently evaluating whether it is possible to extract two factors resembling the conceptual domains of normal and grandiose entitlement. If so, it may be possible to construct an even purer measure of grandiose entitlement. We also recommend that future researchers attempt to develop more items for the NPI-EE (see Ackerman et al. 2012). However, researchers should be careful to add additional items that are content valid and that do not contaminate the scale with other closely related but nevertheless distinct constructs assessing feelings of normal entitlement. It may be helpful for researchers to write additional items that tap into exploitativeness and the derogation of others in the service of promoting the self. This is at the core of vulnerable entitlement as we understand the construct (see also Lessard et al. 2011). Despite these recommendations, we acknowledge that future work is needed to verify this expression of narcissistic entitlement using self-report measures.

Some readers may be concerned that the inclusion of the exploitativeness item on the NPI-EE scale (i.e., NPI item 13, “I find it easy to manipulate people”) impacted differences observed between this measure and the PES.Footnote 10 We investigated whether removal of this item from the NPI-EE would change the findings reported in this paper (complete tables of results are available from the first author upon request). Although removing this item did not generally alter the main conclusions of this paper, the 3-item scale appeared to be somewhat less “disagreeable” and thus more similar to the PES in the FFM Agreeableness domain. For both conceptual (cf. Lessard et al. 2011) and factor analytic (cf. Ackerman et al. 2011) reasons, however, we believe this item should be part of the NPI-EE scale.

Limitations, Future Directions, and Conclusions

Although the current research has contributed additional data to support the argument that the PES and NPI-EE are assessing somewhat different expressions of entitlement, there are several limitations that should be acknowledged. One limitation concerned the use of self-report measures for the criteria in Study 2. If researchers are to be convinced that the two measures predict genuine differences in behavior, it will be important for future research to assess criteria using a method besides self-report. Another limitation of this research was the strict use of college students. It will be important to evaluate the divergence between these measures in other samples, such as participants in community and clinical settings. Moreover, the fact that our samples were made up primarily of White participants is another limitation. In light of recent research indicating racial differences in narcissistic tendencies (Zeigler-Hill and Wallace 2011), it is possible that the current results may not generalize to non-White participants. Thus, as of right now, we can only provide evidence for such differences in samples of mostly White undergraduate students.

Besides the PES and entitlement-based measures from the NPI, there are other measures of entitlement. Some promising alternatives include the Lessard et al. (2011) scales used in the present research. Another potentially useful measure is the Entitlement-Rage subscale from the Pathological Narcissism Inventory (PNI; Pincus et al. 2009). Given the placement of Entitlement-Rage on the Vulnerability scale of the PNI (Wright et al. 2010), it also appears that this measure captures a vulnerable expression of narcissistic entitlement. We hope that researchers will attempt to gather additional validity evidence for these measures while keeping in mind the conceptual distinctions we have proposed in the Introduction.

In sum, the present research suggests that entitlement—much like narcissism itself—is a complicated and multifaceted construct. Although some theorists have positioned entitlement to occur within the context of grandiosity (Millon 2011), it appears that some forms of entitlement may occur within the context of vulnerability (see also Dickinson and Pincus 2003; and Miller et al. 2011b, p. 1032). In addition, there may be differences between normal and narcissistic entitlement (see Lessard et al. 2011). We believe that such distinctions provide increased insight into the differences between the PES and the NPI-EE. Accordingly, future work should continue to refine the understanding of the various facets of entitlement in order to better understand when such feelings are more closely connected to underlying vulnerabilities and low self-esteem as opposed to grandiosity.