Pornography consumption is unusual among regret-pair dilemmas in that the published evidence points to action-side regret exceeding inaction-side regret at the population level — the inverse of the dominant Gilovich-Medvec pattern in most life decisions. Barna Group’s 2015 study of 1,188 US adults aged 25+ (part of a larger N=3,771 multi-instrument design) found that 9% of the general population are currently trying to stop pornography use, 14% would like to use it less, and 18% would rather not use it at all; 18% report “much guilt” about their use while 54% say it does not really bother them. The 9% currently-trying-to-stop figure is the tightest available proxy for action-side regret — it identifies the slice of users whose behavior is dissonant enough with their preferences that they are actively attempting cessation. No published survey directly asks adults “do you regret your pornography use?”, so this active-cessation rate substitutes for the missing regret item. The inaction side is even less directly measured: no large survey asks abstainers whether they regret abstaining, and pornography does not appear as a top long-term regret domain in Roese and Summerville’s 2005 meta-analysis of 11 regret-ranking datasets. The 3% inaction-side rate is an order-of-magnitude estimate anchored on the absence of any documented regret signal among non-users.
The load-bearing caveat is moral incongruence. Grubbs, Perry, Wilt and Reid’s 2019 meta-analysis in Archives of Sexual Behavior (the canonical PPMI model, PMID 30076491) establishes that self-perceived pornography problems — the cluster that includes attempted cessation, guilt, and distress — are predicted far more strongly by the gap between use and personal beliefs about use than by frequency of use itself. Whelan and Brown’s 2021 cross-sectional study of 942 heterosexual men aged 18-44, published in the Journal of Sexual Medicine, makes the same point through a different lens: pornography use alone showed no significant association with erectile dysfunction, premature ejaculation, or sexual dissatisfaction, while self-perceived addiction uniquely predicted all three. The downstream regret signal is generated by the belief that one’s use is problematic, not by the use itself. The 9% Barna headline conflates a religious-conservative subpopulation with high moral-incongruence distress (Barna’s practicing-Christian subsample reports 19% currently trying to stop, 21% wishing to use less, and 40% wanting to stop entirely) with a non-religious population whose use generates near-zero documented distress. Treating the 9% as a single homogeneous regret rate misses the structural fact that the rate is bimodal.
The directional finding — that abstaining from pornography is essentially never regretted in lifetime-regret data while regular use generates a measurable active-cessation signal — should be held loosely on the magnitude side and held firmly on the direction side. The Gilovich-Medvec long-term pattern that dominates most regret-pair entries (84% of biggest-regrets are inactions) does not appear to apply here, because the missed-opportunity mechanism that drives inaction regret in education, career, and romance does not operate symmetrically for pornography: the abstainer who never tried porn does not retrospectively identify a closed window of self-actualization in the way the abstainer who never went to college or never asked someone out does. For occasional users (13-19% of US adults at the 1-2 times per month level, per Barna) the dilemma is less acute than for the regular-user counterfactual modelled here. Cross-national prevalence of problematic-pornography-use criteria ranges from 3.2% to 16.6% across 42 countries in the International Sex Survey (Bőthe et al. PPCS-6 cut-off), so the population denominator for the regret-equivalent group varies substantially by national religious and cultural context, and the 9%/3% US rates should not be transposed directly elsewhere.