Christians are more self-compassionate than atheists, but also more narcissistic

A new study published in Pastoral Psychology explored the psychological traits of Christians and atheists, focusing on self-compassion, narcissism, and shame. The findings suggest that Christians, on average, report both higher levels of self-compassion and higher levels of grandiose narcissism compared to atheists, while both groups reported similar levels of fragile narcissism and shame. These results offer a nuanced view of how religious affiliation may relate to self-perception, emotional well-being, and personality traits.

The study was conducted by Michael W. Magee, an associate professor of psychology at St. Joseph’s University in Brooklyn. Magee specializes in the study of self-compassion and runs a related public education initiative called SelfCompassionNYC.com. He designed this exploratory project to investigate whether self-compassion and narcissism might manifest differently in religious and nonreligious individuals, given the moral teachings and communal frameworks that often characterize religious life.

“This research is the intersection of my long-held research interests in religious attitudes, specifically anti-atheist prejudice (the focus of my dissertation), and my much newer interest in self-compassion,” Magee explained. “I have a daily meditation and self-compassion practice. And I am also a self-compassion practitioner with trained teacher status with the Center of Mindful Self-Compassion. So it just made sense to take a look to see if there were any observable differences in self-compassion between Christians and atheists, two social groups I’ve conducted a lot of research on.”

Self-compassion has been widely studied for its associations with well-being, reduced psychological distress, and increased resilience. However, little research had directly compared self-compassion across religious and secular populations. Magee also included two types of narcissism—grandiose and fragile—and the emotion of shame to explore their relationship with religious identity. Prior work had linked fragile narcissism with shame and had shown that self-compassion interventions often reduce both traits.

To conduct the research, Magee recruited a sample of 631 English-speaking adults from the online platform Prolific. Participants ranged in age from 18 to 40, with nearly equal numbers identifying as Christians (303 people) and atheists (328 people). The sample was predominantly White and came mostly from the United States, the United Kingdom, and Canada. Participants completed a series of validated questionnaires measuring trait self-compassion, grandiose narcissism, fragile narcissism, and shame.

Free daily newsletter

Self-compassion was measured using the Self-Compassion Scale – Short Form, which includes statements such as “I try to be understanding and patient toward those aspects of my personality I don’t like” and “I’m disapproving and judgmental about my own flaws.” Participants rated how often these statements reflected their typical behavior. Grandiose narcissism was measured with the Narcissistic Personality Inventory-13, a questionnaire assessing attitudes such as self-importance, entitlement, and attention-seeking.

Fragile narcissism, sometimes called covert or vulnerable narcissism, was assessed using the Hypersensitive Narcissism Scale, which includes items reflecting self-consciousness and sensitivity to criticism. Shame was measured using the Experience of Shame Scale, which captures self-conscious emotions about behavior, personal attributes, and body image.

The results showed that Christians reported higher levels of self-compassion overall, with statistically significant differences across all six subcomponents of the self-compassion scale. These subcomponents included self-kindness, mindfulness, and common humanity, as well as negative tendencies like self-judgment, isolation, and overidentification. Although it may seem contradictory to score high on both self-kindness and self-judgment, the pattern could reflect the dual role of religious teachings, which often promote forgiveness alongside moral self-scrutiny.

Christians also scored higher on measures of grandiose narcissism, though the differences were modest. The difference was driven primarily by one aspect of grandiose narcissism—grandiose exhibitionism—which includes tendencies to show off and seek admiration. No significant differences were found between Christians and atheists in the other two aspects of grandiose narcissism: leadership/authority and entitlement/exploitativeness.

“My study reveals that religious and non-religious identity shapes self-perception in complex and sometimes unexpected ways,” Magee told PsyPost. “Compared to the atheists in the sample, Christians reported higher levels of self-compassion—being kind and understanding toward oneself—but also showed greater grandiose narcissism, reflecting feelings of moral or spiritual superiority.”

In contrast, no group differences emerged for fragile narcissism. Christians and atheists reported similar levels of hypersensitivity to criticism, self-consciousness, and social anxiety. Likewise, both groups reported similar levels of shame across behavioral, characterological, and bodily domains.

“I was actually surprised by the null shame finding,” Magee said. “There were no observable differences in levels of self-reported shame between either group, which is interesting because shame has a robust direct relationship to narcissism—particularly fragile narcissism. While, at the same time, shame has such a reliable indirect relationship to self-compassion that self-compassion has been referred to as the antidote to shame. I found the equal levels of shame observed counterintuitive.”

These findings complicate simplistic narratives about the emotional life of religious and nonreligious individuals. While Christians may benefit from community support and teachings that encourage compassion and forgiveness, these same settings may also encourage self-enhancement in domains tied to moral or spiritual identity. For example, someone might see themselves as especially virtuous or morally superior because of their religious affiliation—a pattern that overlaps with grandiose narcissism.

“The key message is that psychological strengths and vulnerabilities aren’t determined solely by religious affiliation,” Magee said. “Instead, each individual’s emotional well-being arises from their unique combination of beliefs, values, and experiences. Therefore, supporting emotional health effectively might mean adopting tailored approaches—promoting humility, emotional awareness, and self-compassion—that respect the diverse worldviews people hold.”

But as with all research, there are some caveats to consider. The research relied on self-reports, which can be influenced by social desirability and may not capture deeper unconscious patterns. The sample was also limited to predominantly White and Western participants.

“This was a relatively large online sample, with self-reported data,” Magee explained. “Until similar results are replicated, there are a few caveats. Researchers sacrifice a lot of situational and focal control for the ease of collecting online data. To maximize their earnings, online participants are motivated to work as quickly as possible, which gives researchers reasonable concern about the speed/accuracy trade off. Ideally, we want all of our participants to complete our surveys in as similar a situation as possible and focused on the task.”

“But with an online sample, we sacrifice knowing where our participants are when they are completing the survey, as well as whether they were distracted or not. Are the participants reading things carefully? Are the participants paying attention? Are the participants sober? Researchers try to use various tools, like attention checks and software assistance that will help keeping track of multiple responses from the same individual or bots, but ultimately we place a lot of trust on the online platform to screen their panels effectively.”

“Additionally, the significant results had effect sizes ranging from small to medium, which is not surprising given the size of the sample,” Magee continued. “Observable statistical differences like these in self-reported attitudes do not necessarily translate into any appreciable differences in behavior. And it is really important to keep this in mind and not to over-extend any interpretations.”

Despite these limitations, the study provides a foundation for future work exploring how religious and nonreligious belief systems shape emotional life. It also raises important questions for therapists, counselors, and spiritual care providers. For example, Christians who report high levels of self-compassion may still struggle with narcissistic tendencies, especially if they view themselves as morally superior.

“These results are cool and interesting, but I’m an experimentalist at heart!” Magee said. “Next semester, the students in my Social Cognition Lab at St. Joseph’s University and I will explore ways to experimentally manipulate self-compassion through cognitive-focal priming, and then we will assess the efficacy through various reaction-time measurements. Once that is achieved, it would be interesting to see if such an experimental design has a differential effect on Christians and atheists — or any other social groups for that matter!”

“Most people are more compassionate to other people than they are to themselves,” he added. “Self-compassion is simply bringing ourselves into the circle of kindness we have for the other loved ones in our lives — and self-compassion is a skill that improves with practice. For those who have never explored the practice of self-compassion, I would highly recommend it. It can be especially effective for people with a particularly harsh inner critic.”

The study, “Christians Have More Self‑Compassion Than Atheists—But Also More Grandiose (Not Fragile) Narcissism,” was published July 12, 2025.