Keywords

In an interview just before he died, Bob Ebeling gave a chilling account of being haunted with self-condemnation ever since the explosion of the space shuttle Challenger (Berkes, 2016a, January 8) . As an engineer at NASA, Ebeling urged his directors to delay the launch because cold weather could undermine the integrity of the rubber seals on the booster rockets. Seven astronauts died when this potentiality became a reality. Like many people who are haunted by their past, self-condemnation plagued Ebeling for years after the tragedy. Our team, along with others, are working to understand how self-forgiveness can help individuals regain well-being after such events.

Recent conceptualizations of self-forgiveness advance an approach by which offenders accept an appropriate degree of responsibility for the offense (e.g., Griffin et al., 2015; Woodyatt & Wenzel, 2013) and work to repair their self-image through becoming “decreasingly motivated to avoid stimuli associated with the offense, decreasingly motivated to retaliate against the self … , and increasingly motivated to act benevolently towards the self” (Hall & Fincham, 2005, p. 622) . This two-part definition attempts to differentiate self-forgiveness from a moral disengagement process (or pseudo self-forgiveness ) in which offenders persistently transgress while numbing themselves to guilt and shame (see Gilbert 2017; Leach 2017; Woodyatt, Wenzel, & de Vel Palumbo, 2017).

Given concerns about whether self-forgiveness may facilitate moral disengagement, early scholarship has focused intently on evaluating the degree to which self-forgiveness correlates with well-being, including mental health and relationship quality . A recent meta-analysis (Davis et al., 2015) reported that self-forgiveness was moderately related to a variety of well-being outcomes, including psychological well-being, general mental health, depression , and anxiety . However, meta-analyses are only as sound as the studies they include, and this body of research had two key limitations that inhibit our understanding of the link between self-forgiveness and well-being.

First, most studies reviewed by Davis et al. used measures of self-forgivingness (i.e., tendency to forgive across offenses) that focus only on the self-image repair aspect of self-forgiveness, but do not incorporate responsibility. In the present review, we attend closely to the measurement of self-forgiveness and how that may influence the relationship between self-forgiveness and well-being . Second, most studies reviewed by Davis et al. employed cross-sectional, correlational designs. Thus, the results of the meta-analysis did not give appropriate attention to more sophisticated attempts to operationalize a two-part definition of self-forgiveness that involves an interplay between accurate responsibility and repair of self-image over time. In the present review, we highlight studies that used longitudinal, experimental, or other complex designs (e.g., actor-partner independence model). Failure to attend to these two methodological factors—both involving alignment with the two-part definition of self-forgiveness—could lead to misleading results from meta-analyses.

Qualitative Review

We conducted a qualitative literature review of empirical studies of self-forgiveness and well-being. Our goal was to explore how various ways of operationalizing the two-part definition may partially explain why some studies show a stronger or weaker relationship between self-forgiveness and well-being. Accordingly, we note the measurement strategy (e.g., limiting analyses to those with a certain degree of responsibility; Wohl, Pychyl, & Bennett, 2010; or measuring the process of self-forgiveness rather than merely the repair of self-image ; Woodyatt & Wenzel, 2013) and their potential implications for results . As studies accumulate, our hope is that this qualitative approach can provide the foundation for examining such moderators formally in future meta-analyses.

Literature Search and Inclusion Criteria

We used two methods to locate empirical studies. First, on June 15, 2016, we conducted a PsychINFO search using the term [self-forgiv*]. This search yielded over 190 articles. Second, we obtained the list of references from Davis et al. (2015). We included studies that (a) had a measure of self-forgiveness, (b) had a measure of mental health (e.g., depression , suicidal ideation , well-being, life satisfaction, and substance use) or relationship quality , and (c) were published in a peer-reviewed journal. We did not include measures that might be indirectly related to mental health such as shame or guilt. In total, 65 studies met inclusion criteria including over 20 studies published since the meta-analysis. The method and results of these studies are summarized in two tables that are available upon request from the first author.

Results

Overview of Participants

The reviewed studies used a variety of samples. Most studies (n = 34 of 65) used convenience samples (i.e., undergraduates); however, almost as many (n = 31 of 65) targeted specific applied contexts (e.g., substance abuse treatment , Webb, Robinson, & Brower, 2011; couples, Kim, Johnson, & Ripley, 2011; or separated partners , Rohde-Brown & Rudestam, 2011) . Only two studies included dyadic data of relationships (Pelucchi, Paleari, Regalia, & Fincham, 2013; Pelucchi, Paleari, Regalia, & Fincham, 2015) .

Overview of Measures

Most studies (n = 54 of 65) in the current review assessed self-forgiveness as a trait (i.e., self-forgivingness), the degree to which one tends to forgive oneself across a range of offenses. The most commonly used measures were the Heartland Forgiveness Scale (HFS ; Thompson et al., 2005) and the Mauger Forgiveness Scale (MFS; Mauger et al., 1992) . Only 13 of the 65 studies assessed self-forgiveness as a state. The State Self-Forgiveness Scale (SSFS; Wohl, DeShea, & Wahkinney, 2008) was used in four studies, which requires participants to rate items that assess their current feelings, actions, and beliefs about an identified offense. Several studies (n = 4) adapted trait measures of forgiveness to assess self-forgiveness of a specific offense (e.g., Pelucchi et al., 2013; Wohl & Thompson, 2011) .

Some studies recruited (or selected a subsample of) participants in a manner that ensured individuals accepted some degree of responsibility for the offense (e.g., Pelucchi et al., 2015; Wohl et al., 2010). As we discuss results, we note these strategies. Only Fisher and Exline (2006) explored how responsibility was associated with well-being and used a mental health measure within the scope of this review.

Self-Forgiveness and Mental Health

Trait Measures of Self-Forgivingness

A total of 60 studies have assessed the relationship between mental health and self-forgivingness, including six studies published since Davis et al. (2015). Of the 60, no studies found a negative relationship and only one study found a null relationship between self-forgivingness and mental health (Kaye-Tzadok & Davidson-Arad, 2016) . In this study of 100 female survivors of sexual abuse , self-forgivingness correlated with higher resilience , lower post-traumatic symptoms, but was unrelated to post-traumatic growth . Taken together, self-forgivingness was robustly linked to positive mental health across a variety of outcomes, including depression and mood disturbances (e.g., Bryan, Theriault, & Bryan, 2015; Friedman et al., 2010), anxiety (e.g., Macaskill, 2012), and eating disorders (e.g., Watson et al., 2012) .

Perhaps one of the most compelling lines of evidence of the link between self-forgivingness and positive mental health outcomes is the set of studies on suicidal ideation and behaviors . For example, among military veterans , researchers found a moderately negative relationship between a history of suicide attempts and levels of self-forgivingness (Bryan et al., 2015) . In a sample of domestic abuse survivors , self-forgivingness attenuated the relationship between the frequency of abuse and suicidal behavior (Chang, Kahle, Yu, & Hirsch, 2014) . Although these two studies do not allow us to infer causality, they demonstrate a consistent relationship between higher self-forgivingness and lower suicidal thoughts and behaviors.

State Measures of Self-Forgiveness

Eight of 10 studies that assessed self-forgiveness of a specific offense reported a positive relationship between self-forgiveness and mental health. The two studies that reported a negative relationship included measures of addictive behavior and the stages of change (Squires, Sztainert, Gillen, Caouette, & Wohl, 2012; Wohl & Thompson, 2011) . In Wohl and Thompson , 181 college students trying to reduce smoking and who acknowledged smoking was a “transgression against the self” (p. 356) completed measures of self-forgiveness (only self-image repair ; Brown & Phillips, 2005) and smoking behavior. Those higher in state self-forgiveness were more likely to be in the pre-contemplation stage, and therefore less likely be advancing through the stages of change. Similarly, Squires et al. had 110 college students with signs of a gambling addiction and who were attempting to reduce gambling behavior complete measures of self-forgiveness (Brown & Phillips, 2005), gambling symptomology, and readiness to change . Squires et al. found that higher levels of self-forgiveness negatively predicted readiness to change. Findings from both cross-sectional studies are consistent with the idea that self-forgiveness (specifically, the ability to repair one’s self-image soon after the offense) is associated with a pre-contemplative stage of change, which involves ambivalence about taking the necessary steps required to change one’s behavior.

There were three studies that found a positive relationship between self-forgiveness and behavioral change (Ianni, Hart, Hibbard, & Carroll, 2010; Scherer, Worthington, Hook, & Campana, 2011; Wohl et al., 2010) . In Wohl et al., undergraduates (N = 134) completed state measures of self-forgiveness (adapted from Wohl et al., 2008), procrastination , and negative affect in multiple waves including before and after a midterm . Students were asked whether the procrastination affected their performance with a single three-point item, and any student who replied “not at all” was removed from the study. Results of a mediated-moderation model suggested that, among those who procrastinated on the first exam, self-forgiveness for the offense of procrastination reduced negative affect and made them less likely to procrastinate on a future exam. One way to make sense of the inconsistency between this study and the two described in the prior paragraph is to view the method of dropping participants as a crude way of incorporating the two-part definition of self-forgiveness: each study assessed responsibility differently and various scaling ranges were utilized (e.g., dichotomous versus three-points).

Another important line of evidence comes from two intervention studies (Peterson et al., 2017; Scherer et al., 2011) . Both interventions included content focused on promoting responsibility although neither measured it. Scherer et al. randomly assigned 70 adults diagnosed with alcohol dependence or abuse to a psychoeducational self-forgiveness group or to a control group using treatment as usual. The treatment group reported higher self-forgiveness and self-efficacy to refuse alcohol relative to the control group. Peterson et al. randomly assigned 462 undergraduates who reported an alcohol-related transgression to a self-forgiveness intervention or a neutral condition involving a reflection. Self-forgiveness was moderately and positively associated with an intention to reduce future drinking . Taken together, results from these two interventions are consistent with the idea that self-image repair , if it occurs too quickly and without taking adequate responsibility, can interfere with motivation to change, but as time passes, self-image repair shows a generally positive relationship with mental health outcomes.

Only one study (Fisher & Exline, 2006) assessed how responsibility influences well-being and found no direct relationship. Notably, this study did not report any analysis attempting to incorporate a two-part definition of self-forgiveness, such as examining the link between self-image repair and well-being controlling for responsibility or examining responsibility as a moderator. Nevertheless, the study found responsibility showed a link to outcomes commonly associated with well-being, such as guilt (instead of shame) and remorse (instead of self-condemnation). These findings suggested that accepting responsibility may be indirectly linked to well-being through an emotional coping strategy rather than directly associated.

Taken together, although most studies found a positive relationship between forgiveness of a specific offense and mental health, there were several notable exceptions. These exceptions involved studies that focused on a mental health outcome associated with desire to change a problematic behavior rather than just variables that may correspond to repair of one’s self-image (e.g., psychological well-being). In the one study that attempted explore how responsibility affects mental health, Fisher and Exline showed no direct link between accepting responsibility and well-being and did not incorporate a two-part definition in the analysis.

Self-Forgiveness and Relationships

Trait Self-Forgivingness and Interpersonal Relationships

Of seven studies on self-forgivingness and relationship outcomes, five reported a positive relationship and two reported a null relationship. Of the five, trait self-forgivingness was positively and moderately related to perceived social support in three studies (Day & Maltby, 2005; Jacinto, 2010; Webb, Hirsch, Conway-Williams, & Brewer, 2013). Hill and Allemand (2010) found a positive, but weak relationship between self-forgivingness and the positive relations aspect of psychological well-being (Ryff, 1989), which assesses the number of close friendships and how individuals feel they are perceived. One study (Webb et al., 2011) found a small, positive correlation between self-forgivingness and social support that disappeared over the course of treatment.

The two studies that reported a null relationship used well-being measures focused on relationship quality rather than perceived support. Kim et al. (2011) found a null relationship between self-forgivingness (Thompson et al., 2005) and self-reports of martial satisfaction. Maltby, Macaskill, and Day (2001) found that self-forgivingness (using a single item) was unrelated to indicators of atypical social functioning. Taken together, self-forgivingness showed a consistent, positive relationship with measures of perceived social support , but null effects were more common in studies on self-forgivingness and relationship quality. Thus, perhaps self-forgivingness tends to correspond with perceptions of support, but its actual influence on relationships is more complex and depends on a variety of factors associated with the victim, the offender, and their relationship with each other over time.

State Self-Forgiveness and Interpersonal Relationships

How self-forgiveness of actual offenses affects relationships is largely uncharted territory. Only two studies examined self-forgiveness within relationship dyads (Pelucchi et al., 2013; Pelucchi et al., 2015). In Pelucchi et al. (2013), 168 couples recalled an offense committed against their partner and completed measures of forgiveness (adapted from the HFS ) and relationship satisfaction . If both participants did not accept sufficient responsibility (as measured by a score of three or lower on a seven-point scale) the couple was excluded from the study. The actor-partner model was used to simultaneously estimate both partners’ perspectives of forgiveness and relationship quality . For the offender, higher levels of self-forgiveness and lower levels of unforgiveness correlated with relationship satisfaction ; however, for the victim, only the offender’s unforgiveness of self was associated with low levels of satisfaction. These findings suggest that offenders who persistently experience unforgiveness towards themselves can sour both partners’ view of the relationship over time. Additionally, the positive aspects of self-forgiveness are important for the offender’s, but not the victim’s, sense of satisfaction. In Pelucchi et al. (2015), 130 couples recalled an offense and completed measures of self-forgiveness, relationship quality, and other-forgiveness. They tested a model in which, controlling for severity, self-forgiveness (and unforgiveness) predicted other-forgiveness (and unforgiveness), which in turn predicted relationship quality.

Taken together, there is limited evidence regarding how self-forgiveness affects one’s interpersonal relationships. Self-forgivingness was generally related to perceiving that one has supportive interpersonal relationships, which leads to satisfaction. Notably, only two studies of the seven even included potential offenders and their victims, but these studies did not focus both partners on the same offense, and it is difficult to tell how selecting a subsample based on responsibility might have influenced the results.

General Discussion

The purpose of this review was to examine whether self-forgiveness is associated with benefits to well-being and relationships. Practitioners and scholars have worried that people who learn to repair their self-image too easily and quickly, without appropriate ownership of their hurtful behavior, might promote habits of moral disengagement that could cause great damage to the individual and others (Woodyatt & Wenzel, 2013) . A recent meta-analysis reported that self-forgiveness was moderately and positively related to mental health, but inconsistently related to relationship variables (Davis et al., 2015). In the current chapter, we reexamined these conclusions while considering the various limitations in the studies comprising that meta-analysis. Namely, these studies aligned poorly with a two-part definition that includes an interplay between taking responsibility and repairing self-image . Therefore, we conducted a qualitative review of studies that examined the relationship between self-forgiveness and mental health or relationship quality . Our focus was especially on studies that attempted to incorporate the two-part definition of self-forgiveness through (a) examining forgiveness of specific offenses and (b) incorporating both self-image repair and appropriate responsibility.

Does Self-Forgiveness Promote Mental Health?

As expected, studies that measured self-forgivingness (specifically, self-image repair) were robustly linked with greater mental health (i.e., 59 of 60 studies). In contrast, when mental health and forgiveness was assessed regarding a specific offense, including addictive behavior, results were mixed. Self-image repair did not tend to predict better mental health in studies that focused on change of problematic behavior rather than constructs that conceptually overlap with self-image repair (e.g., psychological well-being). Perhaps the real puzzle is why studies that focused on trait self-image repair (without accounting for responsibility) so consistently predicted well-being. Does moral disengagement largely account for this finding?

Research on sociometer theory (Leary, Tambor, Terdal, & Downs, 1995) would temper such a conclusion. Sociometer theory suggests that self-esteem helps people regulate social acceptance in relationships. Accordingly, people who sever the connection between their reputation with others and their own sense of self would soon become socially isolated, which would tend to damage well-being. This theory suggests that taking responsibility is a long-term strategy for protecting a positive self-image in the face of inevitable transgressions that occur in relationships. In the moment, accepting appropriate responsibility causes moral emotions that may decrease well-being, but the decision to sever relationships is risky and, if used too easily, may result in social rejection that severely undermines one’s ability to maintain high self-esteem. Indeed, several studies (e.g., Griffin et al., 2016; Woodyatt & Wenzel, 2013) testing a two-part definition of self-forgiveness provide indirect evidence (i.e., using measures of guilt, shame, or self-esteem rather than well-being) for the hypothesis that self-forgiveness promotes well-being. However, sociometer theory highlights a gap in the empirical research on self-forgiveness and well-being. Specifically, researchers have not explored the process through which offenders decide whether to repair their relationship with a specific victim or distance from that relationship and seek to protect their broader social reputation through adversarial strategies (e.g., attacking the reputation of the victim).

Although a few studies included responsibility as a moderator of the relationship between self-image repair and well-being, it will be helpful to develop more sophisticated ways of integrating a two-part definition. For example, in scholarship on perfectionism , latent class methods are used to identify categories based on the degree to which individuals have high standards and are critical towards themselves. A similar method could be applied to integrate the two aspects of self-forgiveness. We are especially interested in the possibility that various configurations (i.e., high, low, or medium responsibility) may have strengths and weaknesses for well-being or relationships, depending on the nature of the particular relationship (e.g., degree of exploitation or relationship value).

However, responsibility may be a double-edged sword. In the trauma literature, attribution of responsibility is a major focus of study (Alexander, Eyerman, Giesen, Smelser, & Sztompka, 2004; Janoff-Bulman, 1979) . Individuals often blame themselves for traumatic events and may engage in over-control (i.e., take on too much responsibility), which increases negative outcomes, such as demoralization and depressive symptoms (Janoff-Bulman, 1979) . Therefore, it is with caution one must approach addressing what “appropriate” responsibility is and how to measure it. It is possible that in several of the studies outlined above, appropriate responsibility (i.e., a four or higher) may even be excessive and harmful and could explain the mixed results seen in the addiction studies.

Given the need for greater complexity in basic research on the relationship between self-forgiveness and well-being, emerging intervention work provides an important body of evidence. Initial interventions have showed increases in self-forgiveness and other benefits to mental health (e.g., Cornish & Wade, 2015) . As this work develops, we encourage scholars to draw on theory regarding the regulation and adaptive use of negative emotions (Carver & Scheier, 1998) . On one hand, acute negative emotion can provide a powerful motivator for change, but on the other, chronic negative emotions narrows focus and deplete creativity and coping resources. As demonstrated by Wohl et al. (2010), even early in the process, self-forgiveness may have an important role in reducing rumination and negative emotions. Thus, a productive course of self-forgiveness will likely include the ability to tolerate the negative emotions that come through owning one’s behavior and integrating it into a positive self-image .

Does Self-Forgiveness Promote Better Relationships?

Although only nine studies examined the relationship between self-forgiveness and relationship quality , we can draw a few tentative conclusions. Self-forgivingness correlated positively with perceived social support , and it correlated weakly and inconsistently with relationship quality. The two studies that examined self-forgiveness within actual relationships found that unforgiveness was negatively associated with relationship satisfaction (Pelucchi et al., 2013; Pelucchi et al., 2015). Both studies restricted the sample to those who accepted a certain degree of responsibility for the offense, and it will be helpful to explore this potential moderator with more precise measurement.

Altogether we have more questions than answers about how self-forgiveness affects interpersonal relationships. Most likely, the benefits of self-forgiveness for the offender and others depend on various aspects of the relationship context. Longitudinal studies that track the two aspects of self-forgiveness, personal well-being, and relationship quality over time could help clarify the costs and benefits of various types of forgiveness process. For example, researchers could use latent growth curve modeling to classify people into groups based on their trajectories on measures of self-forgiveness. This approach might clarify how responsibility and other contextual factors, such as a lack of forgiveness from others, might affect relational well-being. We suspect that the ideal process includes an offender who seeks to repair the relationship, accepts responsibility, and then repairs their self-image . This pattern might be associated with better outcomes relative to a trajectory where the offender either uses self-forgiveness to morally disengage or persists in a state of negative emotions.

Strengths, Limitations, and Future Research

There are several notable strengths in this developing literature. First, research on self-forgiveness and well-being has led to the development of a theoretically complex, multi-method, and methodologically diversified field. Second, the potential exists for a thriving applied field of study that can ground and inform basic research. Several studies have already demonstrated positive effects of self-forgiveness interventions (Cornish & Wade, 2015; Griffin et al., 2015; Scherer et al., 2011) . The results of our review suggest that self-forgiveness interventions might be fruitfully extended to other areas, such as treating depression or suicidal ideation or couples counseling.

Despite these strengths, we want to bring attention to several limitations that must be addressed for research in this area to thrive. First and foremost, we documented a major weakness in how researchers are currently attempting to measure a two-part definition of self-forgiveness. The vast majority of studies assessing the link between well-being and self-forgiveness ignore this distinction. Studies that do attempt to ensure that participants have accepted appropriate responsibility either have not used mental health outcomes or have used potentially problematic strategies such as measuring responsibility and then conducting an analysis only on individuals that meet some arbitrary threshold of accepting responsibility. This strategy is tantamount to treating responsibility as a moderator, but not actually comparing the relationship between self-forgiveness and the outcome variable at different levels of responsibility. In the present review, the most common method for incorporating responsibility was using a single item to drop participants based on an arbitrary cutoff. This strategy also forces an assumption that self-forgiveness can only occur after the offender has accepted adequate responsibility for a wrong-doing. Invariably, offenders, victims, and bystanders will have different perspectives of what constitutes “enough” responsibility (Zechmeister & Romero, 2002) . In order to advance our understanding of the benefits of self-forgiveness, we suggest that it is crucial to develop more flexible way of understanding and measuring the responsibility aspect of self-forgiveness.

Within scholarship on self-forgiveness, the typical focus has been on the possibility that people may accept too little responsibility (i.e., moral disengagement ). However, in light of theory and research on trauma recovery, we should be equally concerned that some individuals may practice over-control that causes them to habitually take too much responsibility for offenses. Consider a survivor of domestic violence, someone who lost a spouse during a car accident, or a veteran who was ordered to bomb a community . Does self-forgiveness ever involve a process of decreasing one’s sense of responsibility? In real life, people may encounter offenses that are highly complex and involve conflicting values (e.g., obedience to authority versus a moral code that it is wrong to kill someone). Repairing one’s self-image may sometimes involve creating a new narrative about the offense that involves attenuating or reframing one’s sense of responsibility. If psychologists hope to use interventions to help real people forgive themselves for complex offenses, then they need to fill in the theory on what it means to accept appropriate responsibility for an offense (and to evaluate interventions, we need measures that can capture this process).

Our example at the outset illustrates this issue. Ebeling perceived an offense that haunted him for much of his adult life. Consider what it might have looked like on measures of responsibility and self-image if Ebeling had attended a self-forgiveness intervention that promoted complete healing. Perhaps the intervention would have help him realize, as has been seen in the trauma literature, that he was taking too much responsibility for decisions that he did not make, and his accusations of himself were not realistic or healthy. This insight might have removed barriers to repairing his self-image. Ironically, based on the most common method in the present review, if Ebeling’s score on a responsibility item changed from a 5 (I am very responsible for what happened) to a 1 (What happened was not my fault), then the researcher might have excluded him from the analysis. We believe it is important for future theory and research on self-forgiveness to include the full range of offenses, including those in which self-image repair may require individuals realizing that they are being much too hard on themselves due to perfectionism or over-control coping.

Something similar actually happened for Ebeling, albeit without intervention. Shortly before Ebeling passed away in March of 2016, he found his self-condemnation lessened (Berkes, 2016b, February 25) . After his initial interview, former colleagues reached out to him and emphasized the effort he had expended to halt the launch and reminded him that the decision to launch was outside of his control. According to his family, these conversations stirred an internal shift and his burden grew lighter. In real life, many of the people who seek self-forgiveness may need to reduce the degree to which they feel responsible for what happened. Many of these individuals may have perfectionistic tendencies and live in relatively graceless systems that train and reward high performance (e.g., medicine, athletics, military). The construct of self-forgiveness is too limited if it cannot help these people as well.