Japan banned indoor smoking in restaurants - but a third of pubs and bars still allow it

When Japan's Revised Health Promotion Law took full effect in April 2020, it was supposed to clear the smoke from the country's restaurants, cafes, and bars. Two years later, roughly 70% of restaurants and cafes had gone smoke-free. But walk into an izakaya - a Japanese-style pub - and the odds flip: only about a third are smoke-free. Bars fare even worse, at 25%.

Those numbers come from a new study in Public Health that used data from a major online restaurant review platform to measure the real-world impact of Japan's national smoking ban and the stricter local ordinances layered on top of it in Tokyo and Chiba City.

The loophole problem

Japan's national law came with a built-in escape hatch. Pre-existing small-scale restaurants and bars were temporarily exempted from the ban, allowed to continue permitting indoor smoking as long as patrons under 20 were not present. The logic was pragmatic - forcing thousands of tiny owner-operated establishments to renovate overnight would have been economically disruptive. But the public health consequence is clear: a substantial portion of the hospitality sector remains smoky.

Recognizing this, the Tokyo Metropolitan Government and Chiba City enacted their own stricter passive smoking prevention ordinances, enforced simultaneously with the national law. The research team, led by Professor Isao Muraki at the University of Tsukuba, wanted to disentangle the effects: how much did the national law accomplish on its own, and how much did the local ordinances add?

Parsing the data from 2016 to 2022

Using an interrupted time-series analysis - a statistical method that compares trends before and after a policy change - the researchers tracked the proportion of smoke-free establishments on a restaurant review platform across Japan from 2016 through December 2022.

Nationally, the proportion of smoke-free establishments jumped by an estimated 5.7 percentage points immediately after the Revised Health Promotion Law took effect. In areas covered by the Tokyo and Chiba ordinances, the increase was more than double that: 13.5 percentage points. Of that larger jump, 7.8 percentage points were attributable specifically to the local ordinances rather than the national law.

The breakdown by establishment type tells the fuller story. As of December 2022:

• Restaurants: 68.3% smoke-free
• Cafes: 70.2% smoke-free
• Izakaya (Japanese pubs): 32.8% smoke-free
• Bars: 25.0% smoke-free

Why pubs and bars lag behind

The gap is not mysterious. Izakaya and bars are disproportionately small, owner-operated, and precisely the kind of establishment that qualifies for the exemption. Drinking culture and smoking culture overlap heavily in these venues, and owners may fear losing regular customers. The exemption was designed as temporary, but the study suggests it has created a durable pocket of secondhand smoke exposure.

The researchers emphasize that the findings are not just about numbers on a review platform. Secondhand smoke exposure in enclosed hospitality settings is a well-documented health risk, associated with cardiovascular disease, respiratory illness, and lung cancer in nonsmokers. Workers in these establishments - often young, part-time, and with limited bargaining power - bear a disproportionate share of that exposure.

The case for closing exemptions

The study's conclusion is straightforward: the national law and local ordinances moved the needle, but not far enough. Eliminating the existing exemptions and strengthening enforcement of current regulations would be the most direct path to reducing secondhand smoke exposure across the hospitality sector.

There is an international precedent for this. Countries that implemented comprehensive smoking bans without exemptions - Ireland in 2004, England in 2007 - saw faster and more complete shifts toward smoke-free hospitality. Japan's incremental approach produced incremental results.

Caveats to consider

The study relies on self-reported smoking policies listed on a restaurant review platform, which may not perfectly reflect actual conditions. Establishments that do not list their smoking policy were excluded, potentially introducing selection bias. The interrupted time-series design can account for pre-existing trends but cannot fully rule out confounding from other simultaneous changes - notably, the COVID-19 pandemic, which overlapped with the law's enforcement and may have independently altered restaurant operations and ventilation practices.

The study was supported by Health Labour Sciences Research Grants from the Japanese Ministry of Health, Labour and Welfare, and a medical research grant from the Osaka Association of Adult Diseases Prevention.