Why I dine outside

I’ve switched to dining outside at restaurants. This story from South Korea helps to explain why:

After a woman with the coronavirus visited a Starbucks cafe north of Seoul this month, more than two dozen patrons tested positive days later. But the four face mask-wearing employees escaped infection.

The Aug. 8 outbreak in the South Korean city of Paju is another example of how rapidly the SARS-CoV-2 virus can spread in confined, indoor spaces — as well as ways to minimize transmission. With health authorities around the world still debating the evidence around face masks, the 27-person cluster linked to the air-conditioned coffee outlet adds more support for their mandatory use to help limit the spread of the Covid-19-causing virus.

When I go inside a restaurant or store, I wear a mask. You can’t eat with a mask on, so I eat outside.

PS. Overall, South Korea has a fairly low rate of infection. Thus there’s a good chance that all 27 of those Starbucks patrons got infected at the same place and the same time. And yet not one of the 4 employees got infected.



8 Responses to “Why I dine outside”

  1. Gravatar of mbka mbka
    25. August 2020 at 19:00


    that’s one of the most important lessons and even applied to the original SARS: indoors, air-conditioned spaces are a real and large risk factor (and indoors spaces generally in temperate winters). Mask wearing can mitigate this but outdoors is better yet. During the original SARS outbreak in Singapore I asked architects about the a/c problem. They told me, basically in most office buildings 80% of the air is recycled, 20% is added fresh. Nothing can be done to fix this because it’s by design.

    I am pretty sure that the odd patterns in SE and E Asia, with Thailan, Vietnam, Cambodia etc reporting low rates of infection, but HK, Singapore, South Korea having more trouble (on average) can be explained by absence of presence of a/c. Not to mention the meatpacking plants where perfectly preserved fresh and cold virus is blown by cooling fans at high speeds all over the employees. This problem seems obvious at common sense levels… but hey.

    BTW I never understood the limitations on outdoors activities during the various lockdowns either. Even in Wuhan near the clinics, no outdoor transmission was ever significant.

  2. Gravatar of ssumner ssumner
    25. August 2020 at 19:12

    mbka, Interesting, but I would have thought that cities like Bangkok are now rich enough to have widespread AC. Is that wrong? For the countryside I completely agree.

  3. Gravatar of mbka mbka
    25. August 2020 at 22:36


    maybe I’m wrong on this, as I don’t have actual hard data on a/c saturation. Informally though the degree to which fully seald-off a/c spaces are ubiquitous in Singapore, even Bangkok seems still a bit off. Plus, al fresco dining in Singapore is strictly a thing of the government-organized food courts (“hawker centres”). Literally everything else (or say… 95%?) is indoors and air conditioned in terms of dining.

  4. Gravatar of BC BC
    25. August 2020 at 23:47

    Agree that outdoor dining is safer than indoors and that masks are probably helpful. Those are my priors before reading the article. Most likely, the 27 people did get infected from the Starbucks. But, how many of them were wearing masks, perhaps just picking up a takeout order? How many patrons visited that Starbucks at the same time but did *not* get infected? Also, did the uninfected employees spend most of their time behind the counter, spatially separated from the infected customers and perhaps even separated by a plexiglass barrier? Finally, was there any spatial separation between infected and uninfected customers?

    Article is written to create an impression that 27 out of 27 unmasked individuals got infected while 0 out of 4 masked did not. But what actually happened was that 27 out of 27 infected persons were infected — that’s tautological — with masking ambiguous and 0 out of 4 masked employees were not infected. Uninfected customers are not mentioned.

    The article claims that this story “adds more support for [masks’] mandatory use”. Why wouldn’t it instead “add more support” to a claim that businesses do a pretty good job of protecting employees from infection even in an environment with many infectious customers or that unmasked individuals primarily put themselves at risk without increasing much the risk to masked individuals, which would actually weaken the case for *mandatory* mask usage? If an author does not consider alternative claims that also fit the observations, then what should we conclude about prevailing biases in the way that data is interpreted?

  5. Gravatar of ssumner ssumner
    26. August 2020 at 08:25

    mbka, Thanks for that info.

    BC, I’m against government mask mandates, so I don’t quite see what you are getting at with your “bias” claim.

  6. Gravatar of Gordon Gordon
    26. August 2020 at 13:01

    Even long before this Starbucks superspreader event, there was a case early in the pandemic in a Guangzhou restaurant in which a presymptomatic person infected others up to 12 feet away.

  7. Gravatar of WilliamWykeham WilliamWykeham
    27. August 2020 at 06:35

    @ssumner – why are you against government mask mandates?

  8. Gravatar of ssumner ssumner
    27. August 2020 at 12:11

    William, I don’t wear a mask outdoors, as I believe the risk is low. I do wear a mask indoors (and encourage others to do so), but mandates on indoor wearing should be done by the property owner, for standard Coase Theorem reasons.

Leave a Reply