On average, a person in an abusive relationship attempts to leave seven times before permanently exiting, according to the National Domestic Violence Hotline — a statistic that encapsulates the short-term ambivalence, not a long-term endorsement of staying. An analyst estimate drawn from trauma-bonding and grief literature places action-side regret at roughly 15-20% in the first 12 months after permanent departure — not a rate any single study has measured directly, but one consistent with the short-term regret that trauma bonding and grief over the relationship can produce among a meaningful minority of leavers. The period immediately after separation carries elevated physical risk as well; Campbell et al. (2003), in an 11-city case-control study of intimate partner femicide, found that estrangement — especially from a controlling partner — was itself a preincident risk factor for being killed. These documented costs do not alter the long-term trajectory, but they are real experiences that a clinically honest accounting must register.
The inaction side carries a heavier long-term burden. CDC NISVS 2015 data document that female IPV victims are about three times as likely as male victims to be physically injured (41.6% vs. 13.9%). A 2024 systematic review and meta-analysis in Trauma, Violence, & Abuse (White et al.) found that IPV exposure raises the odds of depression (OR 2.04-3.14), PTSD (OR 2.15-2.66), and suicidality (OR 2.17-5.52) among women. Neither source measures regret directly: the ~55% inaction figure here is an explicit harm-burden proxy synthesized from these injury- and mental-health-prevalence findings, not a survey item asking whether survivors wished they had left sooner.
The 37-point gap between action and inaction regret makes this one of the more asymmetric entries in the dataset, placing it firmly in the inaction-dominates category under Gilovich and Medvec’s temporal framework. Both regret measures are proxies rather than direct bilateral survey items from a common instrument. The central interpretive challenge is that the two sides are not symmetric decision paths: the physical danger during and after separation means that the action carries short-term risks that compound the emotional ones, while the inaction carries long-term psychological and physical harms that accumulate rather than attenuate. The data document a pattern; they do not resolve the individual risk calculation.