Seventy percent of drinkers regretted getting drunk at least once in the prior year, according to the Global Drug Survey 2020 analysis by Davies et al. in Substance Use & Misuse — the largest international study of drinking regret on record (82,821 respondents, 31 countries). That figure is the inaction-side proxy shown above, used because no study directly asks “do you regret continuing to drink alcohol given that non-alcoholic alternatives exist?” head-on. The GDS sample is self-selected, younger, and heavier-drinking than the general population, so 70% is an upper bound on any-regret prevalence among all adult drinkers. The decision framed by this entry is structurally distinct from full sobriety: it is a SWAP, keeping the social ritual, the volume, and the occasions while replacing the alcohol content with 0.0% beer, NA wine, or mocktails.
On the action side, about a third of non-alcoholic-beverage adopters did not report cutting back in the Stanford-summarized Addiction 2024 survey of NA-drink users (Bowdring et al.) — the complement of the roughly two-thirds who said the beverages helped them drink less. We use that ~33% non-reduction share as the action-side swap-ineffective proxy: it captures self-reported non-reduction, not regret about trying, and that survey’s self-reported reduction did not correlate with measured drinking severity. The efficacy anchor is Yoshimoto et al.’s 12-week BMC Medicine 2023 randomized trial — 123 Japanese adults who drink heavily but without diagnosed alcohol dependence, given free NA beverages every four weeks — whose intervention arm reduced alcohol consumption by ~321 g over twelve weeks versus ~77 g in controls (p<0.001), though it reports only group means, not a per-participant non-responder fraction. NielsenIQ’s 2024 market data reinforces that the real-world swap is rarely clean — 93% of US NA-beverage buyers also buy alcoholic beer, wine, or spirits, treating NA products as complements (designated-driver nights, moderation tools, specific occasions) rather than replacements. Stanford Medicine’s 2024 coverage of an Addiction-journal consumer survey reports ~67% of NA adopters with unhealthy-alcohol-use risk said the beverages helped them cut back.
The two figures come from different instruments and populations (Japanese RCT vs global self-selected drinkers) and measure different constructs (intervention non-response vs episodic drinking regret), so the apparent 2:1 inaction-to-action ratio overstates the precision of the cross-side comparison. The directional finding — that continuing to drink alcohol generates substantially more regret than swapping in non-alcoholic alternatives — is robust; the precise magnitude is bracketed by proxies. Notes on alcohol-use disorder: clinical concern exists that the sensory cues of 0.0% beer can trigger cravings for some recovering individuals, but published evidence is mixed and we make no specific claim here. Readers considering the swap for harm-reduction reasons should treat the ~32% reduction signal in Yoshimoto’s intervention arm (a ~321 g drop from a ~996 g/4-week baseline) as a credible real intervention effect; readers considering it for full sobriety should note that the alcohol-vs-abstain entry frames the cleaner decision.