Japanese government study finds silent school lunches were pointless

What Japan’s Covid response lacked in authoritarian nastiness, it made up for in innovative stupidity. And Japan’s experts clearly put much effort into applying their breathtakingly mediocre intellects to the vexing question of how to “stop the spread” during mealtimes. For adults, their solution was mask dining.

Unless you’ve seen people voluntarily practice mask dining in real life, you have no idea how low your opinion of humanity can go.

The other great contribution Japan made to mealtime non-pharmaceutical interventions (NPIs) was silent eating (黙食, mokushoku).

Fortunately, mask dining and silent eating were never widely adopted in restaurants and bars, showing there were limits to the idiocy that most Japanese adults were willing to go along with. But unfortunately, these limits didn’t extend to the idiocy that Japanese adults were willing to make children go along with, so Japanese school kids were forced to eat in silence for 2.5 to 3 years.

But now, at least 3.5 years too late, the Japanese government think tank Research Institute of Economy, Trade and Industry (RIETI) has released a paper estimating the effects of mokushoku titled “The Effect of Silent Eating during Lunchtime at Schools on the COVID-19 Outbreaks”.

Silent eating—no talking during mealtimes—was used as a measure to reduce the spread of infectious diseases during the COVID-19 pandemic because the emission of droplets during conversations was considered a risk factor for spreading the virus. Japan implemented silent eating during school lunchtimes in February 2020, and it remained in effect until November 2022. However, concerns have been raised regarding the potentially negative effects of the policy on children's well-being and educational attainment. More importantly, no study to date has examined its effectiveness in reducing the risk of COVID-19 outbreaks. This study aims to address this important knowledge gap by examining the impact of silent eating on the risk of COVID-19 outbreaks.

So dear reader, are you as excited as I am that this “important knowledge gap” is finally going to be filled?

In November 2022, the Japanese government announced that silent eating in public schools was no longer needed, triggering some schools to discontinue this measure while other schools continued its implementation.

If I may be slightly pedantic, the national government actually announced that they never requested mokushoku in the first place, as I moaned about at the time.

But this is hierarchical and conformist Japan, so rather than angrily demand to know why the government didn’t mention this for 2.5 years, most schools continued mokushoku until March 2023. Anyway, back to the study.

Utilizing this cancelation of the silent eating requirement as a natural experiment, we investigated whether silent eating was associated with a reduced risk of COVID-19 outbreaks. We measured the probability of class closures in public schools (the government’s guidelines required class closures when more than one child in a class was infected with COVID-19) by applying a Difference-in-Differences model with two-way fixed effects to panel data.

Specifically, they looked at data from 11 cities in Chiba Prefecture where 22.3% of schools (45/202) lifted the mokushoku requirement. These schools make up the control group, and the mokushoku schools make up the treatment group. In a Difference-in-Differences model, if a treatment has an effect, the two groups should diverge after an event (here the lifting of mandated mokushoku). So what were the results?

Figure 3 presents the event study figures for the class closures at the classroom level (Panel A), the class closure rate at the school level (Panel B), and the number of class closures at the school level (Panel C). The x-axis shows the days relative to the first day of lifting silent eating in control schools, labeled as day 0.

Event study using De Chaisemartin and d'Haultfoeuille (2020). Panel A shows the results for class closures at the classroom level. Panel B displays the results for class closure rates at the school level, while Panel C is the number of class closures at the school level. The figures plot the average treatment effects for each day, and the bars represent 95% confidence intervals. Standard errors are clustered at the class/school level.

The 95% confidence intervals cross 0 for the entire period after mokushoku is lifted in the control schools, meaning mokushoku didn’t significantly reduce school or class closures for even a single day.

[W]e did not find strong evidence supporting the notion that silent eating during lunchtime at school would decrease the spread of infection.

Say it ain’t so! Next you’re going to tell me partitions don’t work either.

So what’s the takeaway from the filling of this “important knowledge gap”?

Our findings indicate that policymakers should be cautious about using silent eating at schools as a potential lever to control outbreaks of infectious diseases.

I find it ironic that they advise policymakers to “be cautious” since the zero-risk mentality of the Japanese was a major reason mask wearing, silent eating in schools, chanting bans at sports games, and other semi-superstitious behaviour went on for so long here.

I’ve read more papers about NPIs over the last few years than most people could be forced to do at gunpoint, but I’ve never read a paper that managed to be both so unnecessary and so vital at the same time. It’s unnecessary because of course mokushoku doesn’t work! They might as well say that other things policymakers should be cautious about using as potential levers to control outbreaks of infectious diseases include animal sacrifice, self-flagellation, and female genital mutilation. But it’s vital because in future virus scares, anti-mokushoku parents will now have a study to point to, from a Japanese government think tank no less.

But now they’ve released a study showing that talking while unmasked didn’t make a difference during lunchtimes, it’d be nice if the researchers did a study showing that talking while unmasked didn’t make a difference during the rest of the school day either. But when it comes to the pointlessness of masks, Japanese researchers continue to stay silent.