EV-0018
What actually makes forgiveness possible
Fehr, R., Gelfand, M. J., & Nag, M. (2010). Psychological Bulletin, 136(5), 894–914 · Source
What this entitles us to say
A meta-analysis of 175 studies and 26,006 participants finding that forgiveness depends more on the situation than on the forgiver's disposition — with apology, perceived intent, empathy and anger among the strongest correlates.
What it found
175 studies, 26,006 participants, examining forgiveness of a single offender by a single victim. The correlates were grouped into the victim's cognitions, affect and constraints, each with situational and dispositional components.
Notable effects reported: intent r = −.49 (harm that looks deliberate is much harder to forgive), state empathy r = .51, apology r = .42, state anger r = −.41.
The headline structural finding: situational constructs accounted for more variance in forgiveness than victim dispositions did.
Why it matters here
The book's claim is that mercy gets you more. This is the entry that says how it is actually produced.
Two things fall straight out of it into the teaching. First, apology is not a formality — it is one of the larger measured levers on whether forgiveness happens at all, which makes a leader's willingness to apologize a practical skill rather than a matter of pride. Second, perceived intent dominates: the same harm is forgiven or not depending on whether it looked deliberate. That is a direct instruction to leaders about explaining themselves, and it links this trait to the informational dimension in EV-0015.
The situational-over-dispositional finding also protects the first principle. If forgiveness were mostly a personality trait, a low score would be a verdict about who someone is. The research says it is mostly about conditions — which is a spot to work on, exactly as the program claims.
The limits
This is interpersonal forgiveness between individuals, not workplace mercy from a person in authority — the power difference is not modeled, and a leader forgiving a subordinate is not the same act as one peer forgiving another. Correlational. A published correction (2011) revised several reported figures for trait forgiveness, though the authors state the substantive interpretation was unaffected - and note that the correction concerned TRAIT forgiveness figures, not the four coefficients used here. FULL TEXT READ 26 Aug 2026 - 175 studies, 26,006 participants, intent r = -.49, state empathy r = .51, apology r = .42, state anger r = -.41, and situational constructs accounting for more variance than dispositions, all confirmed against the paper.
What we do not say
that these findings transfer cleanly to a boss forgiving an employee. This literature studies peers, and the power difference is not in it. Nor should the coefficients be printed before the full text and the 2011 correction have both been read.