Narcissism and Newlywed Marriage: Partner Characteristics and Marital Trajectories

Abstract

Despite narcissism’s relation with interpersonal dysfunction, surprisingly little empirical research has been devoted to understanding narcissism’s effect on intimate relationships in general or marital relationships in particular. The current study addressed this gap using longitudinal data from a community sample of 146 newlywed couples assessed 6 times over the first four years of marriage. First, we examined partner characteristics associated with higher levels of narcissism to determine the degree to which couples were matched on narcissism and related traits. Second, we examined how narcissism predicted the trajectory of marital quality over time, testing narcissism’s association with initial levels of relationship functioning (the intercept) and changes in relationship functioning (the slope). Results indicated a small degree of homophily but otherwise no clear pattern of partner characteristics for individuals higher in narcissism. Hierarchical linear modeling indicated that wives’ total narcissism and entitlement/ exploitativeness scores predicted the slope of marital quality over time, including steeper declines in marital satisfaction and steeper increases in marital problems. Husbands’ narcissism scores generally had few effects on their own marital quality or that of their wives. These findings are notable in indicating that the effects of personality characteristics on marital functioning may take some time to manifest themselves, even if they were present from early in the marriage. Future research into the mediating psychological and interpersonal processes that link wives’ narcissism with poorer marital functioning over time would be valuable.

Narcissism and Newlywed Marriage: Partner Characteristics and Marital Trajectories

By definition, narcissism is characterized by some level of interpersonal dysfunction, including an exaggerated sense of self-importance and antagonistic behaviors (American Psychiatric Association, 2013). This dysfunction is especially relevant for intimate relationships. Narcissism has been linked to a number of behaviors that could interfere with romantic relationships, including vengefulness (Brown, 2004), domineering and vindictive behavior (Ogrodniczuk, Piper, Joyce, Steinberg, & Duggal, 2009), and interpersonal aggression (Reidy, Foster, & Zeichner, 2010), and narcissistic personality disorder is associated with causing distress in a significant other (Miller, Campbell, & Pilkonis, 2007). Whereas the high level of antagonism (e.g., callousness, exploitativeness) associated with narcissism may help to explain some of this dysfunction, other theories have been put forth to help explain the low commitment seen in narcissistic relationships. For example, Foster, Shrira, and Campbell (2006) showed that the low levels of relationship commitment that characterize narcissistic individuals is linked to overvaluing agentic aspects of relationships (e.g., physical enjoyment) and undervaluing communal aspects (e.g., emotional connections). Nonetheless, some marriage data have revealed that narcissism predicts higher satisfaction and commitment, but only in cases where narcissistic individuals report high self-esteem (Sedikides, Rudich, Gregg, Kumashiro, & Rusbult, 2004) and communal feelings for the partner (Finkel, Campbell, Buffardi, Kumashiro, & Rusbult, 2009).

Despite this work highlighting associations between narcissism and relationship dysfunction, critical gaps remain in our understanding of narcissism’s effect on intimate relationships, particularly marital relationships. The current study addresses these gaps using a longitudinal sample of newlywed couples, providing new insights into the correlates and consequences of narcissism early in couples’ marital trajectories and as they unfold over time.

What Characterizes the Partners of Narcissistic Individuals?

One critical question relates to whether the individuals whom narcissistic individuals pursue, date, and marry have specific personality profiles. Two competing hypotheses arise from the existing literature. In general, the relationship literature finds evidence for homophily effects, such that similar people tend to become romantically involved (e.g., Decuyper, De Bolle, & De Fruyt, 2012). Given these findings, along with other work indicating that individuals higher in narcissism value self-oriented qualities (e.g., ambition, confidence) more highly than caring qualities (e.g., considerate) in their ideal romantic partner (Campbell, 1999), narcissistic individuals may partner with other narcissistic individuals. Alternatively, given the antagonistic behaviors associated with narcissism, the opposite might be true: narcissistic individuals might seek partners who are pliable and willing and able to bend to the needs and desires associated with narcissistic individuals’ self-absorption. There is no clear evidence for this hypothesis, though many resources exist online and in book form that are addressed to “victims” or “prey” of the ex-partners of narcissistic individuals.

Only recently have researchers collected empirical data on the psychological characteristics of individuals who have or are currently dating narcissistic individuals and the limited evidence available suggests that there is no clear pattern of partner type. Three dyadic studies have found modest evidence of homophily (Campbell, Foster, & Finkel, 2002, Study 5; Keller, Blincoe, Gilbert, DeWall, Haak, & Widiger, 2014; Lamkin, Campbell, vanDellen, & Miller, 2015). There has been no supported evidence of the “victim hypothesis” thus far, although a recent study did find a negative trend against assortative mating for narcissism (Smith et al., 2014). These studies primarily included dating couples, however, so our ability to draw conclusions about the types of partners that narcissistic individuals actually marry is limited.

Narcissism and the Trajectory of Relationship Satisfaction over Time

In addition to the question of whom narcissistic individuals marry, it is unclear if and at what point narcissistic traits give rise to relationship dysfunction. Prior work offers conflicting hypotheses. These studies have generally investigated two components of analytic models: the intercept, reflecting initial level of functioning, and slope, which indicates changes in functioning over time. Personality characteristics such as neuroticism and depressive symptoms have been shown to be associated with the intercept of marital satisfaction rather than its slope (Karney & Bradbury, 1997; Kurdek, 1998), consistent with the enduring dynamics model of relationship functioning (Huston, Caughlin, Houts, Smith, & George, 2001) which argues that couples’ difficulties arise early in their relationships and remain consistent over time. Accordingly, this view would suggest that traits such as those associated with narcissism will be related to initial levels of marital satisfaction and should predict more negative initial levels (i.e., the intercept).

Alternatively, there is good reason to believe that, in the case of narcissism, relationship dysfunction may only emerge over time. Paulhus (1998) observed that strangers perceive narcissistic individuals positively upon initial meeting, but their perceptions grow negative over time. Extending this finding to romantic relationships, Campbell (2005) theorized that partners of narcissistic individuals are likely to rate their relationship satisfaction especially highly early in their relationship, but are likely to endorse lower satisfaction later in the relationship. This has been called the chocolate cake model, in that chocolate cake is initially enjoyable to eat but would become unpleasant or have unpleasant consequences if consumed in large quantities over time. For the narcissistic individual himself/herself, a cost-benefit framework may help to explain how relationship satisfaction unfolds (Campbell & Campbell, 2009). There are few costs, or risks, to initially engaging in a relationship for a narcissistic individual. It may bring benefits such as opportunities for self-enhancement (e.g., associating with influential others to boost one’s self esteem) and requires little investment. However, a long-term relationship is associated with effortful costs for narcissistic individuals (e.g., commitment, fidelity), resulting in lower satisfaction as time passes. Taken together, these perspectives suggest that marital functioning for narcissists and their partners should start high but then worsen significantly more than is typical over time. Some evidence supports this assertion, but only via cross-sectional data thus far. In dating relationships where some aspects of grandiose narcissism were present in both members of the couple, relationships of a longer duration had lower satisfaction than relationships in earlier stages (Lamkin et al., 2015).

Limitations of the Existing Literature

Our ability to answer these foundational questions about the partners of narcissistic individuals and how their marriages unfold is limited in several ways. First, research about outcomes of narcissism in relationships has been conducted primarily using retrospective or cross-sectional data. There is a need for prospective, longitudinal data, ideally early in couples’ marital trajectories, in order to determine whether relationship dysfunction is present from early in couples’ relationships or whether it emerges over time. Specifically, given the conditional temporal impact of narcissism, it is important for new studies to include methods of analysis that investigate change in functioning over time (i.e., slopes) in addition to initial level (i.e., intercepts; Karney & Bradbury, 1995).

Second, few studies have taken a dyadic approach to understanding how narcissism manifests in relationships, as much of the literature focuses on only one member of the couple. The limited research employing dyadic data has used student samples of dating couples, thereby providing potentially questionable generalizability regarding how narcissism impacts partners in more committed relationships. Dyadic studies of narcissism’s relations with romantic functioning examined in community samples of married couples would provide a more nuanced look at narcissism’s effects on more established relationships, especially as the commitment required for marriage is more formal and public than ties related to dating, which can be more easily broken if satisfaction declines.

Another methodological limitation concerns the fact that many studies examining the role of narcissism have used the total score from the Narcissistic Personality Inventory (NPI; Raskin & Terry, 1988), which does not allow for more fine-grained identification of the behaviors and qualities that drive relationship difficulties. In particular, some researchers have identified a subscale that is strongly associated with negative outcomes (entitlement/exploitativeness), while other subscales (leadership/authority and grandiose/exhibitionism) have more mixed characteristics and are linked to both adaptive and maladaptive outcomes (Ackerman et al., 2011). Examining subscales from the NPI allows for the possibility that certain aspects of narcissism may promote more positive relationship outcomes, others may not affect relationship outcomes, and others may promote more negative relationship outcomes. For example, entitlement/exploitativeness has been shown to be linked with narcissistic rivalry, which in turn has been associated with more revenge-focused reactions in response to relationship transgressions and lower levels of general empathy, trust, forgiveness, and gratitude (Back et al., 2013), all of which are likely to lead to poor relationship functioning. Conversely, the more purely “grandiose” components of the NPI such as leadership/authority and grandiose exhibitionism (Miller et al., 2014) may have more mixed relations to outcomes given the positive (i.e., extraverted, assertive) and negative (antagonistic, aggressive) behaviors associated with these traits (e.g., Küfner, Nestler, & Back, 2013). In addition, although narcissism is often associated with strategies to enhance self-esteem, there is a distinction between self-esteem and narcissism (e.g., Sedikides et al., 2004). However, some have asserted that the NPI includes items that assess psychologically healthy self-esteem (Rosenthal & Hooley, 2010). For this reason, it is helpful for studies to establish that patterns that emerge for narcissism (especially for certain subscales) are not confounded by relations with self-esteem.

Lastly, much of this literature has not distinguished between male and female partners in relationships. It is widely acknowledged in dyadic research that there are important differences (e.g., Kiecolt-Glaser & Newton, 2001), with wives generally seen as the “barometers” of marital functioning in heterosexual marriages (Floyd & Markman, 1983). There are also known gender differences in narcissism. In a meta-analysis, Grijalva and colleagues (2014) demonstrated that males had higher narcissism scores on the NPI than females, and this was driven by subscale score differences in entitlement/exploitativeness and leadership/authority. Thus, it is important to consider gender when answering questions about narcissism in romantic relationships.

The Present Study

Given the increased interest and implications of narcissism’s effect on romantic relationships, we investigated partner characteristics and marital functioning over time in a longitudinal, community sample of newlywed married couples that was significantly larger than those used in past research on narcissism in marriage. The design of the sample has several notable strengths. First, the early years of marriage are an ideal time to study these associations, as marital satisfaction declines on average (e.g., Lavner & Bradbury, 2010) and risk of marital dissolution is high (e.g., Kreider & Ellis, 2011). The newlywed years also mark a unique period after couples have already made a serious commitment to one another but before their relationship dynamics have become too well-established, thus allowing for the possibility to predict changes in functioning, not just overall level of functioning. Second, the sample includes data from both spouses, which provides: (1) a more accurate portrayal of partner characteristics (e.g., the partner’s own ratings versus the narcissistic individual’s ratings of the partner), and (2) the possibility of examining the effects of one partner’s level of narcissism on his/her own marital trajectory as well as on the partner’s trajectory. Including data from husbands and wives also allows us to consider unique patterns of association based on gender.

Our first aim was to examine partner characteristics associated with narcissism. We hypothesized that some evidence of homophily would be present for narcissism itself. We also tested the alternate “victim” hypothesis by examining whether narcissism was associated with specific partner Big Five personality profiles. For these analyses, evidence of homophily would be manifested via correlations between one partner’s narcissism scores and the other’s scores on traits of extraversion (positive) and agreeableness (negative; Campbell & Miller, 2013). Conversely, evidence of the victim hypothesis would receive some support if partners of narcissistic individuals were high on traits associated with victimization and psychopathology, such as neuroticism and introversion (e.g., Kotov, Gamez, Schmidt, & Watson, 2010; Rogosch & Cicchetti, 2004; Slee & Rigby, 1993) or agreeableness (Gore & Pincus, 2013).

Our second aim was to examine how narcissism was associated with the trajectory of marital functioning over time, including partners’ reports of their marital satisfaction as well as their reports of their marital problems. Consistent with the chocolate cake and cost-benefit models outlined by Campbell and colleagues (2005, 2009), we hypothesized that narcissism would predict better intercepts (i.e., initial levels) but worse slopes (i.e., degree of change) for marital satisfaction and marital problems (e.g., greater decreases in satisfaction and greater increases in marital problems). We tested this against the rival hypothesis from the broader marital literature showing that maladaptive personality characteristics typically predict lower intercepts of marital quality. In doing so we examined overall narcissism as well as its subscales, and also included self-esteem to ensure that any observed effects were specific to narcissism. We also examined within-spouse (e.g., wives’ narcissism to wives’ satisfaction) and cross-spouse (e.g., wives’ narcissism to husbands’ satisfaction) effects to determine whether any effects for narcissism are seen for the narcissistic individual, the partner, or both members of the couple.

Method

Participants

Participants were drawn from a larger study of 169 newlywed couples that began in 2001 in a Northern Florida community surrounding a major state university.¹ Couples were recruited by (a) placing advertisements in community newspapers and bridal shops, offering payment to couples willing to participate in a study of newlyweds and by (b) sending invitations to eligible couples who had completed marriage license applications in counties near study locations. Couples responding to either solicitation were screened for eligibility in a telephone interview. Inclusion required that this was the first marriage for each partner; the couple had been married less than 6 months; each partner was at least 18 years of age; each partner spoke English and had completed at least 10 years of education (to ensure comprehension of the questionnaires); couples did not have children; and wives were not older than 35. Eligible couples, after providing oral consent, were scheduled for an initial laboratory session.

Husbands averaged 25.6 (SD = 4.1) years of age and 16.3 (SD = 2.4) years of education; 59% were employed full time, 34% were full-time students, and 94% were White. Wives averaged 23.4 years of age (SD = 3.6) and 16.2 (SD = 2.0) years of education; 45% were employed full time, 45% were full-time students, and 86% were White.

Procedure

Couples meeting eligibility requirements were scheduled to attend a 3-hour laboratory session within the first 6 months of marriage. Before the session, participants were mailed a packet of questionnaires to complete at home and bring with them to their appointment, along with a letter instructing couples to complete all questionnaires independently of one another. Upon arriving to the session, spouses completed a written consent form approved by the local human subjects review board, participated in a variety of tasks beyond the scope of the present study, and were paid $70.

At approximately 6-month intervals subsequent to the initial assessment, couples were recontacted by telephone and again mailed questionnaires, along with postage-paid return envelopes and a letter of instruction reminding partners to complete forms independently. This procedure was used at all follow-up procedures except at Time 5, which resembled Time 1 in that couples completed questionnaires at home and brought them to the laboratory, where they engaged in a variety of tasks beyond the scope of this study. After completing each phase, couples were mailed a $40–50 check for participating.

Narcissism was assessed at Time 3. Accordingly, we set Time 3 as the initial assessment (hereafter Wave 1 in the text and tables) and examined marital trajectories from this point forward, thereby providing a total of six assessments of marital satisfaction and marital problems from years 1.5 to 4.5 in marriage (i.e., Waves 1–6). One hundred forty-six couples provided narcissism data and are included in the following analyses.

Measures

Marital satisfaction

Marital satisfaction was assessed at Waves 1–6 using the Quality of Marriage Index (QMI; Norton, 1983), a six-item scale asking spouses to report the extent to which they agree or disagree with general statements about their marriage (e.g., “We have a good marriage”). Five items ask spouses to respond according to a 7-point scale, whereas one item asks spouses to respond according to a 10-point scale, yielding scores from 6 to 45. Higher scores reflected greater satisfaction. Coefficient alpha was > .90 for husbands and for wives across all phases of the study.

Marital problems

Spouses’ marital problems were assessed at Waves 1–6 using a modified version of the Marital Problems Inventory (Geiss & O’Leary, 1981). This measure lists 19 potential problem areas in a marriage (e.g., in-laws, sex, showing affection) and asks participants to “indicate how much it is a source of difficulty or disagreement for you and your spouse” on a scale from 1 (not a problem) to 11 (major problem). We summed specific problem ratings into an overall index of problem severity that could range from 19 to 209. This measure had high internal consistency (α > .85 for husbands and wives across all assessments).

Marital satisfaction and marital problems were significantly correlated (median cross-sectional |r| = .64 for husbands and .69 for wives). Nonetheless, we examine each measure separately given (a) conceptual distinctions between marital satisfaction and marital problems (Fincham & Bradbury, 1987) and (b) previous work showing that marital satisfaction can change even as marital problems remain stable (Lavner, Karney, & Bradbury, 2014).

Narcissism

Both partners completed the Narcissistic Personality Inventory (NPI; Raskin & Terry, 1988), a 40-item, forced-choice assessment of grandiose narcissism. Their score on this measure at Wave 1 was used as the independent variable in all analyses described below. In addition to the total narcissism score (husband α = .82; wife α = .82), we used three subscales based on work by Ackerman and colleagues (2011): leadership/authority (LA – 11 items; husband α = .76; wife α = .67), grandiose exhibitionism (GE – 10 items; husband α = .63; wife α = .67), and entitlement/exploitativeness (EE – 4 items; husband α = .41; wife α = .39). The subscales were significantly correlated, with effects in the moderate range (median |r| = .32 for husbands and .36 for wives, see Table 2, Panel 1). Six-month test-retest correlations for husbands in the current sample (n = 127) were 0.79 for total narcissism, 0.79 for leadership/authority, 0.73 for grandiose exhibitionism, and 0.62 for entitlement/exploitativeness. For wives (n = 129), six-month test-retest correlations were 0.75 for total narcissism, 0.78 for leadership/authority, 0.76 for grandiose exhibitionism, and 0.56 for entitlement/exploitativeness.

Table 2.

Within- and Cross-Spouse Correlations for Husbands’ and Wives’ Total Narcissism and Subscale Scores(N = 145 couples)

Panel A: Within-Spouse Correlations
	NPI TOT	NPI LA	NPI GE	NPI EE
Husbands
NPI TOT	—
NPI LA	0.85**	—
NPI GE	0.65**	0.37**	—
NPI EE	0.56**	0.36**	0.22**	—
Wives
NPI TOT	—
NPI LA	0.80**	—
NPI GE	0.69**	0.33**	—
NPI EE	0.66**	0.46**	0.30**	—
Panel B: Cross-Spouse Correlations
	Wives
	NPI TOT	NPI LA	NPI GE	NPI EE
Husbands
NPI TOT	0.13	0.17*	0.04	0.02
NPI LA	0.07	0.10	−0.01	−0.02
NPI GE	0.07	0.10	0.10	−0.08
NPI EE	0.11	0.14	−0.05	0.07

^*p < .05.

^**p < .01.

Notes. TOT = total narcissism. LA = leadership/authority. GE = grandiose exhibitionism. EE = entitlement/exploitativeness.

Means and standard deviations for marital satisfaction, marital problems, and narcissism are shown in Table 1. Consistent with other research on newlyweds (e.g., Kurdek, 1998; Lavner & Bradbury, 2010), the sample on average had fairly high levels of marital satisfaction and low levels of relationship problems. Levels of total narcissism were also consistent with prior research (e.g., Twenge, Konrath, Foster, Campbell, & Bushman, 2008).

Table 1.

Means and Standard Deviations for Narcissism, Marital Satisfaction, and Marital Problems

	Husbands	Wives
	Mean (SD)	Mean (SD)
Narcissism (Wave 1)
NPI TOT	18.35 (6.04)	17.48 (5.94)
NPI LA	6.75 (2.71)	6.00 (2.46)
NPI GE	3.02 (1.95)	3.03 (2.00)
NPI EE	1.17 (1.04)	1.37 (0.99)
Marital satisfaction
Wave 1	40.22 (6.20)	41.13 (5.80)
Wave 2	40.85 (4.92)	41.81 (4.60)
Wave 3	40.26 (6.65)	40.49 (5.92)
Wave 4	39.69 (7.14)	40.54 (6.47)
Wave 5	38.66 (7.39)	39.67 (6.52)
Wave 6	39.19 (6.49)	40.31 (6.27)
Marital problems
Wave 1	50.44 (23.77)	49.21 (22.91)
Wave 2	50.75 (22.98)	46.76 (20.64)
Wave 3	51.24 (26.85)	47.49 (23.27)
Wave 4	52.57 (26.19)	50.13 (23.02)
Wave 5	56.32 (29.79)	48.10 (22.38)
Wave 6	49.95 (24.16)	47.75 (21.68)

Notes. TOT = total narcissism. LA = leadership/authority. GE = grandiose exhibitionism. EE = entitlement/exploitativeness.

Big Five personality

We assessed the Big Five model of personality at Wave 1 using the 50-item International Personality Item Pool (IPIP; Goldberg, 1992), a set of personality inventory questions in the public domain. Items are rated on a 1 (strongly disagree) to 5 (strongly agree) scale. Internal consistency was high for each dimension: neuroticism (husband α = .84; wife α = .89), extraversion (husband α = .90; wife α = .89), openness to experience (husband α = .79; wife α = .82), agreeableness (husband α = .83; wife α = .81), and conscientiousness (husband α = .82; wife α = .84).

Self-esteem

Spouses’ self-esteem was assessed at Wave 1 using the 10-item Rosenberg Self-Esteem Questionnaire (Rosenberg, 1965). Scores on this measure can range from 4 to 40, with higher scores indicating higher self-esteem (sample item: “I take a positive attitude toward myself”). Internal consistency was high for husbands and wives (α > .80).

Results

Narcissism and Partner Personality

We began by examining the association between spouses’ narcissism scores and their partner’s personality at Wave 1. First we examined the degree of homophily in partners’ narcissism scores (Table 2, Panel 2). Of the 16 correlations, only 1 was significant: husbands’ total narcissism manifested a small positive correlation with their wives’ leadership/authority score. All other correlations were small and non-significant (median r = .07).

We then examined whether spouses’ narcissism scores were associated with particular partner personality profiles (Table 3). For husbands, only 3 of the 20 correlations between their narcissism and their wives’ Big Five personality traits were significant. Husbands’ total narcissism was negatively associated with their wives’ agreeableness, as was husbands’ leadership/authority. Husbands’ entitlement/exploitativeness was positively associated with their wives’ conscientiousness. For wives, only 2 of the 20 correlations between their narcissism and their husbands’ Big Five personality traits were significant: wives’ grandiose exhibitionism was positively associated with their husbands’ openness, whereas wives’ entitlement/exploitativeness was negatively associated with their husbands’ extraversion. The effect sizes for the five significant correlations between one spouse’s narcissism and the other spouse’s Big Five traits were small in nature (median |r| = .17). All other correlations were non-significant.

Table 3.

Correlations between Spouses’ Narcissism Scores and their Partners’ Personality Characteristics(N = 145 couples)

	Partner Personality
	Openness	Conscientiousness	Extraversion	Agreeableness	Neuroticism	Self-Esteem
Husbands
TOT	0.12	0.07	−0.14	−0.18*	0.03	0.06
LA	0.10	−0.01	−0.13	−0.17*	0.07	0.05
GE	0.03	0.00	−0.09	−0.14	0.03	0.10
EE	−0.01	0.23**	−0.03	−0.05	−0.01	0.07
Wives
TOT	0.06	0.02	−0.02	−0.09	−0.03	0.05
LA	0.04	0.05	0.05	−0.09	−0.02	0.03
GE	0.17*	0.08	−0.01	0.10	−0.13	0.06
EE	−0.01	−0.12	−0.17*	−0.13	0.10	−0.03

^*p < .05.

^**p < .01.

Notes. Correlations represent the association between spouses’ narcissism scores and their partners’ personality characteristics (e.g., husband leadership/authority and wives’ openness). TOT = total narcissism. LA = leadership/authority. GE = grandiose exhibitionism. EE = entitlement/exploitativeness.

Narcissism and Newlyweds’ Marital Trajectories

To examine whether husbands’ and wives’ narcissism scores were associated with trajectories of marital satisfaction and marital problems over the early years of marriage, we used growth curve analytic techniques and the HLM 7.0 computer program (Raudenbush, Bryk, & Congdon, 2010). Growth curve analytic techniques allow for a two-level process in data analysis. Level 1 allows for the estimation of within-subject trajectories of change (growth curve) for a variable, described by two parameters: an intercept (initial level of the variable) and a slope (rate of change over time). Level 2 allows for the examination of between-subjects differences in these parameters using individual-level predictors.

Husbands’ and wives’ data were estimated simultaneously within the same equations (as opposed to nesting spouses within couples) to allow for sex-specific intercepts, slopes, and random effects (Atkins, 2005). Time was estimated as the number of assessments after the baseline assessment and was uncentered so that the intercept terms (B_f00 and B_m00) could be interpreted as the value at baseline (Wave 1). We used the following equations:

$$Y_{\text{ti}}(\text{Outcome})={(\mathit{\text{female}})}_{\text{ti}}[{\mathrm{\pi }}_{\mathrm{f}0\mathrm{i}}+{\mathrm{\pi }}_{\mathrm{f}1\mathrm{i}}{(\mathit{\text{Time}})}_{\text{ti}}]+{(\mathit{\text{male}})}_{\text{ti}}[{\mathrm{\pi }}_{\mathrm{m}0\mathrm{i}}+{\mathrm{\pi }}_{\mathrm{m}1\mathrm{i}}{(\mathit{\text{Time}})}_{\text{ti}}]+{\mathrm{e}}_{\text{tij}}$$ Level 1

$${\mathrm{\pi }}_{\mathrm{f}0\mathrm{i}}(\text{wife intercept})={\mathrm{\beta }}_{\mathrm{f}00}+\text{husband narcissism}+\text{wife narcissism}+{\mathrm{\mu }}_{\mathrm{f}0\mathrm{i}}{\mathrm{\pi }}_{\mathrm{f}1\mathrm{i}}(\text{wife slope})={\mathrm{\beta }}_{\mathrm{f}10}+\text{husband narcissism}+\text{wife narcissism}+{\mathrm{\mu }}_{\mathrm{f}1\mathrm{i}}{\mathrm{\pi }}_{\mathrm{m}0\mathrm{i}}(\text{husband intercept})={\mathrm{\beta }}_{\mathrm{m}00}+\text{husband narcissism}+\text{wife narcissism}+{\mathrm{\mu }}_{\mathrm{m}0\mathrm{i}}{\mathrm{\pi }}_{\mathrm{m}1\mathrm{i}}(\text{husband slope})={\mathrm{\beta }}_{\mathrm{m}10}+\text{husband narcissism}+\text{wife narcissism}+{\mathrm{\mu }}_{\mathrm{m}1\mathrm{i}}$$ Level 2

These equations include separate intercepts and linear slopes for men and women, and sex-specific variance components at Level 2. Husbands’ and wives’ narcissism scores were included simultaneously at Level 2, allowing us to examine how the parameters of spouses’ marital trajectories were affected by their own narcissism and their partner’s narcissism (e.g., husbands’ marital problems predicted by their own narcissism and their wives’ narcissism).

We began by conducting eight separate models, one for each outcome of interest (marital satisfaction and marital problems) and narcissism dimension (total narcissism, leadership/ authority, grandiose exhibitionism, and entitlement/ exploitativeness). Results for total narcissism scores are shown in Table 4. For husbands and wives, wives’ total narcissism was associated with the slope of marital satisfaction and the slope of marital problems, such that marriages in which wives had higher levels of total narcissism were marked by greater declines in satisfaction and greater increases in problems (all p < .05).

Table 4.

Associations between Total Narcissism and Marital Satisfaction and Marital Problems Trajectories

	Intercept			Linear Slope
	Estimate (SE)	t ratio	Effect size r	Estimate (SE)	t ratio	Effect size r
Husbands
Marital satisfaction
Husband TOT	−0.03 (0.09)	−0.29	0.02	0.02 (0.02)	0.87	0.07
Wife TOT	0.08 (0.08)	1.06	0.09	−0.06 (0.02)	−2.79**	0.23
Marital problems
Husband TOT	−0.23 (0.39)	−0.58	0.05	−0.04 (0.07)	−0.56	0.05
Wife TOT	−0.04 (0.34)	−0.12	0.01	0.17 (0.08)	2.20*	0.18
Wives
Marital satisfaction
Husband TOT	−0.04 (0.07)	−0.57	0.05	0.01 (0.02)	0.43	0.04
Wife TOT	0.07 (0.07)	0.91	0.08	−0.04 (0.02)	−2.24*	0.19
Marital problems
Husband TOT	0.52 (0.31)	1.68	0.14	−0.05 (0.07)	−0.70	0.06
Wife TOT	−0.20 (0.28)	−0.73	0.06	0.15 (0.07)	2.23*	0.19

^*p < .05.

^**p < .01.

N = 143 couples. df = 140 for husbands and wives. Effect size r = sqrt [t²/(t² + df)].

We then examined univariate results for the three dimensions of narcissism. Leadership/authority was not associated with the intercept or slope of either of the outcome measures for husbands or for wives (Table 5). Grandiose exhibitionism showed distinct patterns of results for husbands and wives (Table 6). For husbands, the slope of marital satisfaction and the slope of marital problems were predicted by their own and their wives’ scores, such that husbands’ grandiose exhibitionism predicted better marital trajectories (higher satisfaction, decreases in problems), but their wives’ grandiose exhibitionism predicted worse marital trajectories (declining satisfaction, increases in problems). Wives’ initial level of marital problems (i.e., the intercept) was also predicted by their husbands’ grandiose exhibitionism, such that women married to men higher in grandiose exhibitionism reported more problems.

Table 5.

Associations between Leadership/Authority and Marital Satisfaction and Marital Problems Trajectories

	Intercept			Linear Slope
	Estimate (SE)	t ratio	Effect size r	Estimate (SE)	t ratio	Effect size r
Husbands
Marital satisfaction
Husband LA	0.07 (0.20)	0.34	0.03	0.02 (0.04)	0.49	0.04
Wife LA	0.17 (0.22)	0.77	0.06	−0.05 (0.05)	−1.11	0.09
Marital problems
Husband LA	−1.25 (0.82)	−1.52	0.13	0.07 (0.14)	0.48	0.04
Wife LA	−0.22 (0.82)	−0.27	0.02	0.25 (0.18)	1.43	0.12
Wives
Marital satisfaction
Husband LA	−0.11 (0.16)	−0.68	0.06	0.00 (0.04)	0.13	0.01
Wife LA	0.24 (0.22)	1.11	0.09	−0.07 (0.04)	−1.48	0.12
Marital problems
Husband LA	0.55 (0.70)	0.79	0.07	−0.02 (0.14)	−0.12	0.01
Wife LA	−0.86 (0.75)	-1.15	0.10	0.23 (0.16)	1.51	0.13

N = 143 couples. df = 140 for husbands and wives. Effect size r = sqrt [t²/(t² + df)].

Table 6.

Associations between Grandiose Exhibitionism and Marital Satisfaction and Marital Problems Trajectories

	Intercept			Linear Slope
	Estimate (SE)	t ratio	Effect size r	Estimate (SE)	t ratio	Effect size r
Husbands
Marital satisfaction
Husband GE	−0.32 (0.26)	−1.23	0.10	0.11 (0.05)	2.20*	0.18
Wife GE	−0.02 (0.24)	−0.07	0.01	−0.13 (0.06)	−2.07*	0.17
Marital problems
Husband GE	1.19 (1.04)	1.14	0.10	−0.41 (0.19)	−2.17*	0.18
Wife GE	−1.19 (1.02)	−1.17	0.10	0.47 (0.21)	2.18*	0.18
Wives
Marital satisfaction
Husband GE	−0.27 (0.25)	−1.09	0.09	0.03 (0.04)	0.81	0.07
Wife GE	−0.19 (0.20)	−0.94	0.08	−0.05 (0.05)	−1.02	0.09
Marital problems
Husband GE	2.72 (1.11)	2.46*	0.20	−0.18 (0.20)	−0.89	0.08
Wife GE	0.60 (0.87)	0.69	0.06	0.25 (0.19)	1.32	0.11

⁺p < .10.

^*p < .05.

^**p < .01.

N = 143 couples. df = 140 for husbands and wives. Effect size r = sqrt [t²/(t² + df)].

Next we examined the effects of entitlement/exploitativeness (Table 7). Results were identical to those for total narcissism. For husbands and wives, wives’ entitlement/ exploitativeness was associated with the slope of marital satisfaction and marital problems, such that marriages in which wives had higher levels of entitlement/exploitativeness were marked by greater declines in satisfaction and greater increases in problems (all p < .05). As with total narcissism, husbands’ entitlement/exploitativeness did not have a significant effect on marital trajectories for themselves or for their wives.

Table 7.

Associations between Entitlement/Exploitativeness and Marital Satisfaction and Marital Problems Trajectories

	Intercept			Linear Slope
	Estimate (SE)	t ratio	Effect size r	Estimate (SE)	t ratio	Effect size r
Husbands
Marital satisfaction
Husband EE	0.40 (0.47)	0.84	0.07	−0.08 (0.09)	−0.84	0.07
Wife EE	0.05 (0.48)	0.10	0.01	−0.40 (0.14)	−2.82**	0.23
Marital problems
Husband EE	−1.69 (2.00)	−0.84	0.07	−0.02 (0.36)	−0.07	0.01
Wife EE	3.67 (2.14)	1.72	0.14	1.00 (0.50)	2.02*	0.17
Wives
Marital satisfaction
Husband EE	0.42 (0.36)	1.17	0.10	−0.02 (0.08)	−0.28	0.02
Wife EE	−0.05 (0.43)	−0.12	0.01	−0.28 (0.11)	−2.55*	0.21
Marital problems
Husband EE	−1.82 (1.51)	−1.21	0.10	0.19 (0.34)	0.57	0.05
Wife EE	1.11 (1.61)	0.69	0.06	0.83 (0.42)	1.98*	0.17

^*p < .05.

^**p < .01.

N = 143 couples. df = 140 for husbands and wives. Effect size r = sqrt [t²/(t² + df)].

Finally, to see if these results were influenced by self-esteem (Sedikides et al., 2004), we conducted a series of analyses in which we repeated the analyses described above while simultaneously covarying out self-esteem (simultaneously entered at Level 2). We focused only on total narcissism, grandiose exhibitionism, and entitlement/exploitativeness, as there were no significant effects for leadership/authority. Results indicated that the effects for narcissism – notably all of the effects on slopes – remained the same when controlling for self-esteem. Moreover, unlike narcissism, self-esteem was a significant predictor of intercepts, not slopes, and higher levels of self-esteem were associated with higher quality marriages (e.g., higher satisfaction and fewer problems).

Discussion

Narcissism has long been associated with interpersonal dysfunction, but empirical research on how narcissism affects marital relationships is lacking, limiting conceptual understandings of how narcissism manifests within established intimate relationships. Using data from a community sample of 146 newlywed couples assessed 6 times over the first four years of marriage, we addressed two key questions regarding narcissism in the context of marriage.

First, we examined whether narcissistic individuals marry partners with specific characteristics. We hypothesized that the partners of individuals higher in narcissism would demonstrate homophily (e.g., have some narcissistic characteristics, have higher levels of extraversion and lower levels of agreeableness) rather than exhibit characteristics of a “victim” (e.g., higher levels of neuroticism, lower levels of extraversion). However, we did not expect any homophily effects to be particularly strong, based on previous studies (Keller et al., 2014; Lamkin et al., 2015). Overall, more evidence was found for the homophily hypothesis than for the victim hypothesis. The only significant correlation consistent with the victim hypothesis was that women with higher levels of entitlement/exploitativeness were married to men lower in extraversion. No significant associations were found between any of the other subscales of narcissism and extraversion, or for narcissism and partner’s neuroticism. Consistent with the homophily hypothesis, there were some relations between total narcissism, leadership/authority, and agreeableness. More specifically, women with higher levels of leadership/authority were married to men with higher levels of total narcissism. In addition, men higher in total narcissism and leadership/authority were married to wives lower in agreeableness. These data add to a growing literature that suggests that there is only relatively modest evidence to suggest that narcissistic individuals partner with specific types – be they those with similar (i.e., homophily) or dissimilar (i.e., victim hypothesis) traits.

Our second research question was how narcissism affected the trajectory of marital satisfaction and problems over time. We investigated this question using spouses’ reports of their marital satisfaction and marital problems, two variables which we expected to be related but not synonymous (Lavner, Karney, & Bradbury, 2014). We tested two competing models: one drawing from Campbell’s (2005) chocolate cake model in which narcissism would predict better initial levels of marital quality but significantly worse slopes over time, and a second drawing from the broader marital literature showing that maladaptive personality traits generally predict lower intercepts of marital quality (e.g., Karney & Bradbury, 1997). Narcissism was not associated with better initial functioning, as the chocolate cake model would predict, but there was some evidence that narcissism predicted the slope of marital quality. Specifically, wives’ total narcissism predicted worse slopes of marital satisfaction and marital problems for themselves and for their husbands, as did wives’ entitlement/exploitativeness, consistent with the idea that the entitlement/ exploitativeness subscale is strongly associated with negative outcomes (Ackerman et al., 2011). In addition, wives’ grandiose exhibitionism predicted worse marital trajectories for their husbands (but not for themselves), whereas husbands’ grandiose exhibitionism predicted better marital trajectories for themselves. No significant effects were found for leadership/authority on husbands’ or wives’ marital trajectories. The only evidence supporting the hypothesis that narcissism would predict poorer initial levels of marital quality was that husbands’ grandiose exhibitionism predicted greater problem intercepts for their wives. Overall, these findings indicate that the relationship impairment associated with narcissistic traits may take some time to manifest, and highlight the value of studying these associations longitudinally to clarify whether this dysfunction was present from early in marriage or emerged over time.

Before discussing the broader implications of these results, it is important to acknowledge some of the limitations of this study. First, while its longitudinal design is a notable methodological strength, our study still only captures a portion of the relationship – the early years of marriage. This timeframe is commonly studied in the marital literature (e.g., Kurdek, 1998; Lavner & Bradbury, 2010), and is widely seen as a time of significant change for many couples, as newlyweds on average experience declines in satisfaction across the first years of marriage (e.g., Kurdek, 1998) and their risk of marital dissolution is high compared to more established couples (e.g., Kreider & Ellis, 2011). Nonetheless, a key question that remains is: what patterns emerge after the newlywed years? It would be particularly interesting to examine whether the greater declines in satisfaction and increases in problems documented here continue for several years, or whether they eventually level off as spouses settle into more familiar roles and/or adjust their expectations. It would also be valuable to examine these couples even earlier in their marital trajectories, or possibly even before they married, in order to better assess initial differences in relationship quality. Second, although we included data from both members of the couple, all ratings were self-report. Narcissistic individuals see themselves more positively than others see them (e.g., Carlson, Vazire, Oltmanns, 2011), suggesting that these ratings may be underestimates. Including partner reports of personality – particularly partner ratings of narcissism – would provide another valuable perspective. Ratings of partner personality could also allow for interesting tests about partner perceptions, such as whether narcissistic individuals perceive their partners as being more similar to themselves than their partners’ own views. Future research should investigate these questions.

There are also limitations relating to our use of the NPI. Although this is a widely-used measure of narcissism that corresponds well with grandiose and DSM-based variants of the construct (i.e., narcissistic personality disorder) (Miller, Lynam, & Campbell, in press; Miller et al., 2014), it provides less information regarding the more vulnerable components of narcissism. Further research is needed regarding how vulnerable narcissism manifests in the context of intimate relationships such as marriage, as vulnerable narcissism is generally characterized by a quite disparate nomological network (e.g., Miller et al., 2011). Moreover, the entitlement/ exploitativeness subscale demonstrated relatively low reliability, though still consistent with prior research which has similarly indicated negative outcomes associated with this domain (e.g., Ackerman et al., 2011; Lamkin et al., 2015). Even so, some caution is needed when interpreting the findings for this scale. Other assessments of the antagonistic aspects of grandiose narcissism using measures such as the Five-Factor Narcissism Inventory (FFNI; Glover, Miller, Lynam, Crego, & Widiger, 2012) may be useful. Finally, the present study represents a single sample, and further replication with other samples is needed.

Notwithstanding these limitations, these findings have significant implications for theoretical understandings of narcissism in the context of newlywed marriage. Our findings showing relatively low levels of homophily in narcissism between spouses is somewhat inconsistent with findings from the broader marital literature documenting assortative mating in personality (e.g., Decuyper, De Bolle, & De Fruyt, 2012; cf. Luo & Klohnen, 2005). At the same time, we note that very limited evidence was found for the opposite victim hypothesis, in which individuals higher in narcissism pair with partners who might be more pliable and easier to manipulate. These findings build on recent work using a sample of undergraduate dating couples which similarly showed stronger evidence for homophily than for the victim hypothesis (Lamkin et al., 2015). Interestingly, we note that the study of dating couples showed more consistent evidence for homophily. This discrepancy across studies raises the possibility that the types of partners with which narcissistic individuals are willing or able to form less serious dating relationships may differ from marriage relationships – or that narcissistic parings are not stable enough to result in marriage – and underscores the importance of studying romantic partnerships at varying stages of relationship trajectories before drawing general conclusions. Taken together, our findings indicate that narcissistic individuals marry partners who may have some narcissistic features but ultimately do not have a clear “type,” and call for further study of the psychological processes involved in partner selection among individuals high in narcissistic traits.

Results also demonstrated that wives’ narcissism was associated with worse changes in marital quality over time for themselves and for their husbands, but not with initial levels of marital quality. These findings are especially noteworthy when compared to previous work showing that personality characteristics predict the intercept (initial level) of marital quality rather than the slope (e.g., Karney & Bradbury, 1997). Indeed, additional analyses among the current sample indicated that self-esteem was positively associated with the initial level of satisfaction. The fact that narcissism was associated with changes in satisfaction over time indicates that the negative effect of certain personality characteristics may only emerge as time passes, even though these traits were present early in marriage. Whether this differential patterning of results is especially true for more pathological personality characteristics such as narcissism would be a valuable direction for future research.

These findings also underscore the importance of considering total narcissism as well as its component parts. The subscale analyses consistently showed the detrimental effects of the entitlement/exploitativeness dimension for husbands and wives, building on prior work highlighting the multiple impairments associated with this domain (e.g. Ackerman et al., 2011). Moreover, consistent with prior research (Lamkin et al., 2015), the more adaptive forms of narcissism (leadership/authority and grandiose exhibitionism) generally showed few positive or negative effects on relationship functioning, which may be due to its association with both attractive and repelling qualities (e.g., Küfner et al., 2013). This work calls attention to the fact that different facets of grandiose narcissism are likely to yield distinct effects on relationship functioning, and supports the idea that the positive, self-enhancement aspects of narcissism can lead to positive or neutral social outcomes, whereas the negative, antagonistic aspects of narcissism often lead to negative social outcomes (Back et al., 2013).

Further research is needed to explore the mediating psychological and interpersonal processes that explain how narcissistic traits eventually manifest into heightened relationship dysfunction. For example, narcissism has been linked to increased attention to alternatives (i.e., appraising other potential partners while in a current relationship; Campbell & Foster, 2002) and game-playing (i.e., displaying unclear and inconsistent commitment to the relationship; Campbell et al., 2002), which would be expected to contribute to relationship dysfunction. Other researchers have recently focused on describing behaviors commonly associated with narcissism such as aggression (e.g., Reidy et al., 2010). Future research should examine whether these findings generalize to the marital context through increased levels of verbal and physical aggression toward one’s partner.

The gendered nature of these findings is also notable. Sex differences are frequently explored in relationship research (e.g., Kiecolt-Glaser & Newton, 2001) but are somewhat understudied in narcissism research, although broad differences are certainly noted (e.g., males having overall higher narcissism scores than females; Grijalva et al., 2014). In the present study, wives’ characteristics and perspective appeared to drive most of the outcomes observed in our results. One broad interpretation of this pattern is through the lens of marital research, which has shown that wives’ characteristics disproportionately affect the outcome of heterosexual relationships (e.g., Floyd & Markman, 1983). However, there may also be something unique about narcissism itself. For example, to the extent that men are generally characterized as having higher levels of narcissism than are women (Grijalva et al., 2014), there may be something normative about men having higher levels of narcissistic features, making narcissism less deleterious for their romantic functioning, whereas women’s narcissistic features may be seen as especially negative. Additional replication and further exploration of these findings is warranted, as is research with same-sex couples to determine how much of these findings are due to sex versus other relational factors (e.g., in a female same-sex couple, whether both partners’ narcissism affects relationship quality or only the narcissism of one partner).

In conclusion, the data reported here provide new insights into how narcissism is associated with partner characteristics and the trajectory of relationship quality within the context of newlywed marriage. These findings raise new questions about gender differences in the effects of narcissism on marital relationships, and call for further research on how wives’ narcissism proves detrimental to marital functioning over time.