Gradual escape from lockdown: bubble merge

May 4, 2020 10:49 AM

We could see people with reduced infection risk by merging `bubbles' while they're healthy, re-splitting when one of them has a case. That link is a fairly general description of how big our `bubbles' would be, how we would decide to merge and re-split them. It's based on pre-2020 data suggesting that a lot of people spend most of their time in a small geographical region already.
posted by clew (34 comments total)

It's based on pre-2020 data suggesting that a lot of people spend most of their time in a small geographical region already.

Except for the great many who commute to jobs in a different geographical region than the one where we live, and where the act of commuting may put us into close proximity with people who are living in yet a different geographical region and working in still yet another geographical region....
posted by EmpressCallipygos at 10:52 AM on May 4, 2020 [11 favorites]

Yes, but this isn't a plan for pre-pandemic, it's a plan for getting us partly out of lockdown within a pandemic. Local employment, or quarantine after movement, is safer than unlimited movement.
posted by clew at 10:55 AM on May 4, 2020 [4 favorites]

Doesn't this rely on enough testing being available that you can determine if you're an asymptomatic carrier?
posted by sagc at 10:56 AM on May 4, 2020 [12 favorites]

Yeah, this seems completely moot until we have a rapid, accurate test that everyone can give themself every morning when they wake up. Untested asymptomatic people make this plan useless.
posted by hydropsyche at 10:59 AM on May 4, 2020 [15 favorites]

TL;DR: Enforce lockdowns at smaller regional levels instead of national levels AND limit/ban commuting between regions at different lockdown levels.
posted by vacapinta at 11:00 AM on May 4, 2020 [5 favorites]

Which demographics tend to have the healthiest bubbles? Which demographics would have limited freedom of movement longest using this method?
posted by aniola at 11:02 AM on May 4, 2020 [12 favorites]

Which demographics tend to have the healthiest bubbles? Which demographics would have limited freedom of movement longest using this method?

Well, during the Trump Administration, "Votes Republican" will be the only criteria used.

This sounds flippant, but I would bet my last dollar that this is what would happen in the United States of America if the legal parts were worked out (and I doubt they can, honestly) in order to make this bubble stuff happen.
posted by sideshow at 11:06 AM on May 4, 2020 [4 favorites]

There's a natural testable division to use -- your bubble is your sewage system, because we can test sewage for shed coronavirus.
posted by clew at 11:06 AM on May 4, 2020 [18 favorites]

Yes, but this isn't a plan for pre-pandemic, it's a plan for getting us partly out of lockdown within a pandemic. Local employment, or quarantine after movement, is safer than unlimited movement.

I understand that. My point is that the number of people who are not locally employed is not a statistically insignificant one, and it's a surprising oversight in the original study - unless the definition of "local" is really flexible. In fact, I fear that because of all the great many commuters we have, and the great variety of distances in which each commuter travels, we would either spend a good deal of time getting bogged down in what the definition of "Local" is, in an effort to strike the right balance between "only six mom-and-pop shops open up, everyone else is SOL" and "open maybe 50% of the business up but we have 100 people who are commuting in from outside the area, including our CEO, so that creates problems".
posted by EmpressCallipygos at 11:14 AM on May 4, 2020 [2 favorites]

we can test sewage for shed coronavirus

A very useful trick, but lots of people live in rural areas with wells and septic systems. Can we assume that if a metro area is relatively virus-free then the surrounding rural areas are too? Seems like you'd need to start with a lot of rural area individual test results to figure out the correlation.
posted by echo target at 11:19 AM on May 4, 2020 [3 favorites]

unless the definition of "local" is really flexible

Yeah, I live in Orange, CA, and a reasonable "local" to me includes 10 million people. Hell, since I used to fly up to the home base up in the South Bay twice a month, for a time "local" to me was just about all 30m people in California.
posted by sideshow at 11:24 AM on May 4, 2020 [5 favorites]

I found this startling because where I am (BC, Canada) the language of bubbles is being used very differently. A policy currently under discussion is that we each pick one other household of people we care about and form a two-household bubble. Slightly higher R0 (everyone in the bubble is at risk if one person is infected), but only slightly, and with a considerably improved quality of life.

The idea of asking someone to join my bubble seems as fraught as proposing marriage.
posted by justsomebodythatyouusedtoknow at 11:38 AM on May 4, 2020 [39 favorites]

I only get within 50’ of someone when absolutely necessary. Fortunately my job has virtualized pretty well to enable this, but e.g. re-opening Axe-throwing venues, cinemas etc. is lunacy borne of desperation.
posted by Heywood Mogroot III at 12:04 PM on May 4, 2020 [2 favorites]

Meanwhile Australia and New Zealand will discuss forming a travel bubble.
posted by needled at 12:05 PM on May 4, 2020 [3 favorites]

Yes, the conversation here has ben about bubbles being a MUCH smaller group of people, not a region-wide bubble (and the problems that has created including toxic siblings "racing" to physically claim the grandparents FIRST without any discussion is pretty gross). Seems more sensible to start small and move gradually towards larger interconnected groups than to just declare a whole area "open for business".

I am in a tourist area and the amount of out-of-towners coming during the lockdown is way out of control. If an area opens up a lot of people will travel to get their dinner or shopping done, spreading the risk even further.
posted by saucysault at 12:07 PM on May 4, 2020 [4 favorites]

Both these papers cite Tolber and Sizer (1997), and the first link talks about commuting zones "of 5,000 to 100,000 inhabitants". The commuting zones in Tolber and Sizer have an average size of over 300,000 people and the largest is 14,545,373. And that's the 1990 census; the US has only increased population and increased the distance of interactions in the past 30 years. Did they even read the papers they cite?

The definition of an area based on commuting flows is the way metropolitan areas in the US (and other areas, including Canada and the EU) are defined.* A metropolitan statistical area in the US is based on including counties that have at least 25% of their workforce commuting into the metro area. That seems like a fairly plausible level to me. The metro areas are one of the most heavily cited statistical geographies; the list is here, and it starts with New York-Newark-Jersey City, NY-NJ-PA MSA at 19.5 million and Los Angeles-Long Beach-Anaheim, CA MSA at 13.2 million.

If you use the cutoff of 15% commuting, there's a second, larger statistical area definition - the combined statistical area. New York gets Bridgeport, New Haven, Poughkeepsie and Trenton added and is 22.6 million. LA gets the Inland Empire and Ventura and grows to 18.8 million. The entire greater Bay Area CSA including Stockton, Modesto and Santa Cruz is 10 million; so is the combined Washington-Baltimore area. Some of these are less traditionally considered a single area (Boston with Providence and Worcester, Portland and Salem) and some are more traditionally considered a single area (the Wasatch Front - Provo, SLC and Ogden; Raleigh and Durham). But at this level, there's 36 CSAs with 2 million and 25 with 1-2 million.

That's the most available definition of a commuting zone, and they're huge. And there's still substantial commuting between the edges of the regions; Lakeland, FL has a lot of commuting into both Orlando and Tampa so no matter what side of the boundary it's on (it's in the Orlando CSA), a lot of people would get cut off.

What is important is to consider that many of these are multi-state areas, so there needs to be multi-state coordination on policy -- luckily, a lot of that is already happening with the regional agreements already in place.
posted by Homeboy Trouble at 1:05 PM on May 4, 2020 [10 favorites]

I look forward to enjoying telling my friends "Sorry but I actually know you and therefore do not trust you"
posted by srboisvert at 1:36 PM on May 4, 2020 [4 favorites]

How many people have been tested in the US? Do we really even know?Last estimate I saw was we've only tested something like 7 million people. Out of 210 million adults? Seriously?

Any idea we're re-opening the economy is totally delusional.

Until we can test/antibody test every single working adult every few weeks (or a vaccine is developed) we are not going to have a functioning economy without re-starting this and killing a whole lot of people and crashing it al all over again.

And without all that who is going to return to normal activities when there is so much we don't know? When there is so much federal incompetence? Does anybody really believe that gyms and restaurants and retail stores are going to return to anything close to previous business? Who the fuck is going to go to a commercial gym? We're going to trust individual responsibility? We're going to trust the mighty market?

Who the fuck is going to "go back to normal" right now?"

Careless idiots. That's who. The fucking morons who are stomping around with guns demanding FREEDUMB.
posted by Everyone Expects The Spanish Influenza at 3:05 PM on May 4, 2020 [11 favorites]

I was kind of shocked to see maps of Manhattan in that 2nd link, since it seems like the prime example of a place where the bubble model would be difficult to impossible to pull off. According to a 2013 census estimate, under normal circumstances, the population of Manhattan literally doubles during the day as people commute in from surrounding neighborhoods, boroughs, counties, and even states. I am, ordinarily, one of those people.

People may not stray far from work for lunch or drinks, but you probably can't reopen an office if 90% or more of the people who work there, and in all of the surrounding buildings, are commuting in from different places. And as goes people who both live and work in Manhattan, how do you deal with the fact that grocery stores and other essential businesses are only kept up and running by people who commute in?

I can see how this might work in areas that are already relatively contained, but I'm not sure it pans out for heavily affected high-population areas where entire neighborhoods only function because of the labor of people who don't live within them.
posted by evidenceofabsence at 3:53 PM on May 4, 2020 [3 favorites]

TORONTO WAS MADE FOR THIS! We're cliquey as hell. I've been going to parties with the same friends and animation crowd people for 35 years!
posted by bonobothegreat at 6:29 PM on May 4, 2020 [2 favorites]

There are interesting twists on this whole idea, too. This one is sort of inside-out. Everyone I work with comes from within the same area bubble (greater Spokane region). But tonight, we sent a driver over to Missoula MT and neighboring regions to make night deliveries. We also sent a driver up to Kalispell, MT and surrounds on a night delivery run. We also sent someone over to Wenatchee and up to Omak and back through Grand Coulee. And a fourth driver headed out of here down through the Tri-Cities to Pendleton, OR and then looping back through Walla Walla.

The theoretical reach of the bubble I work in covers four states, multiple 400-500 miles driving loops... I don't know how to find the right map to link, but you see the vast region on Google Maps. These deliveries are nearly all made after business hours into unoccupied buildings or dropboxes there for our use. The amount of human contact during these routes is pretty limited. Maybe a gas station attendant or two along the way and that's all aside from the people in the bubble at the warehouse before departing.

But it's that gas station pump that interests me most. Of anything my drivers encounter out on their route that is the object that is mostly likely to have been touched by multiple people from perhaps vast distances apart.

I've been having conversations with my drivers, casually but repeatedly, about being aware of the things they touch and what they touch after that and how they are navigating, specifically, gas stations and gas pump handles.

My personal method is, once I finally found them a couple of weeks ago, to keep a canister of sanitizing wipes with me and pull out one of those and use that to open door handles, handle the gas pump, wipe off the gas selector buttons and the keypad, and to keep bounding that towel around in my hands. I might use a bare finger when I do punch in my driver number, but I wipe the keypad down again when I'm done and that pad is just right there, in my hands, being held, rubbing against my fingers.

But yeah, this idea of what could be brought where some places unknown with an entirely different outbreak experience from here (we've been quite low, really)... I think about what I touch.
posted by hippybear at 7:10 PM on May 4, 2020 [1 favorite]

My personal method is, once I finally found them a couple of weeks ago, to keep a canister of sanitizing wipes with me and pull out one of those and use that to open door handles, handle the gas pump, wipe off the gas selector buttons and the keypad, and to keep bounding that towel around in my hands. I might use a bare finger when I do punch in my driver number, but I wipe the keypad down again when I'm done and that pad is just right there, in my hands, being held, rubbing against my fingers.

Personally, I keep a little rubber tipped tablet stylus in my "outside pants" pocket. Finally a use for the thing. It works really well for elevator buttons, handicap door openers and self-checkout screens (except for Walmart for some reason). I bought it almost a decade ago thinking it would help with tableting and never ever used it. Along comes a pandemic and my hoarder tendencies have been rewarded for another couple of decades.
posted by srboisvert at 7:21 PM on May 4, 2020 [4 favorites]

Consider a town with some 10,000 inhabitants where most people work in a local factory, or in other occupations within the community.

Err.

Okay. I shall so consider. *ponder ponder ponder*

The last time I spent serious time in or near such a town was over 30 years ago, when I lived in rural Arkansas. There are five times that many people in my current zip code, about a 3 mile radius. I cannot even imagine sectioning the neighborhood into 10k people zones.

Once the virus in under control, no new infections have been detected for several consecutive days, the town would be labelled “green”. Its inhabitants would then be able to return to their usual social and economic interactions within their zone but travel out of town would be restricted.

That's...

Okay, there's some validity there, but it's utterly useless in, oh, all the urban centers that are where we're desperately trying to flatten the curve.

And unless we can stop travel between "safe" districts, it's almost entirely meaningless. And I don't expect Ms. "I Want A Haircut" to agree to get one at whatever the closest salon is, instead of her favorite one that's across town or 30 miles away in another city.
posted by ErisLordFreedom at 8:53 PM on May 4, 2020 [9 favorites]

The last time I spent serious time in or near such a town was over 30 years ago, when I lived in rural Arkansas. There are five times that many people in my current zip code, about a 3 mile radius. I cannot even imagine sectioning the neighborhood into 10k people zones.

Same. In the postcode where I currently live in Chicago to have an area with under 10,000 people would be less than 1/3rd of a square mile. I live in the very densest part of this dense neighborhood so it would likely be even smaller than that.
posted by srboisvert at 6:07 AM on May 5, 2020

We can think of this using graph theory.

The perfect solution is that everyone is in a bubble of one person. No further transmission. Of course, we mostly do not live that way so this doesn't work. Instead we form bubbles of household size, prune all our connections outside that bubble to the absolute minimum required.

The next step up is to merge bubbles. Two, maybe three households with a large degree of trust between them. This reconnects the graph, increasing transmission but maybe only a little.

The problem is in figuring out the next step after that. Households and small groups of (probably related) households are "natural" groupings. Easy to understand what is required and not usually that big. Of course there is the challenge that a household can only join one multi-household bubble, once you go beyond that the degree of connection escalates way too fast. In some cases that will work (my wife's family all lives near me, my family and the other inlaw families live in other countries so there is not a hard choice to make), in other cases it will not.

What's next though? We need to find bubble sizes that are big enough to be useful and there are not many logical layers in our society between family and "the whole city" that could be used to configure bubbles (remember, it only works if all the smaller sized bubbles connect to only one larger sized bubble). Neighborhood is not usually an economic unit. Some towns may be relatively self contained but you really need *a lot* of isolation between units to make a difference from a fully connected graph.

This is the theory behind dividing workers into multiple shifts / teams with no contact between them, break the social graph into isolated chunks which are big enough to carry out work but not so big that you reconstruct the whole social graph.

Here in London, as in NYC, commuters make it essentially impossible to re-open much. On the one hand, these service based economies are also strongly polarised in terms of who is currently able to work. In manufacturing based economies, you cannot operate at all when people cannot go to work but when they can, almost everyone can work. In service based economies there are office workers who are able to do substantial parts of their job (often all) from home but also a whole group of restaurant, pub, yoga studio, etc. workers who usually provide services to the former but who cannot do their jobs remotely.
posted by atrazine at 6:50 AM on May 5, 2020 [1 favorite]

I mean, this is a good idea on paper. But it is depending on conditions which represent a smaller percentage of reality than exists in the real world.

I mean, my hometown fits the population part okay, but it's an old New England mill town so the one "local factory" shut down 20 years ago, and today most people commute either to the nearest big cities (Hartford or New Haven), to crap jobs in neighboring towns of similar size, or to the state university two towns over.

And I don't think this is uncommon among most communities - a couple decades' worth of shit economy and decaying manufacturing have greatly decreased the idea of there being the One Big Company In Town Where Everyone Works; more often, you're likely to find towns where there's The One Big Company That Closed Up And Moved To China Or Mexico And Laid Everyone Off 20 Years Ago, so now everyone commutes elsewhere because that's all they could find in the meantime. Where there is "One main industry in town", that "town" has ballooned in size tremendously to support the people flocking there for work.

So....this is a great idea on paper, but it would have worked better if this was, like, 1983.
posted by EmpressCallipygos at 6:53 AM on May 5, 2020 [4 favorites]

The next step up is to merge bubbles.

Why?
posted by Heywood Mogroot III at 7:21 AM on May 5, 2020

Why?

Because people don't like isolating in individual households and because some parts of our society require physical presence outside the household to operate. If there are options that allow control of transmission while allowing those activities to take place they should be considered.
posted by atrazine at 9:02 AM on May 5, 2020 [1 favorite]

Why?

Because my father is dying of cancer and my mother is desperately lonely. And needs more help than hospice is able to provide.
posted by cooker girl at 11:08 AM on May 5, 2020 [3 favorites]

Humans are social creatures by nature, and the alternative to discussion on what constitutes a reasonable expansion of the bubble (and also which bubble we're talking about) is people will just start ignoring the rules anyway and making decisions based on the philosophical principle known as YOLO. Shouting people down and calling them dumb isn't going to have the results you're looking for.

There's room for a by-the-numbers analysis on what a bubble merge means, that doesn't involve licking door handles and a second wave and people needlessly dying. Part of what's necessary is a total rethinking of what constitutes a risky activity. Risky activities under pandemic rules: going shopping at the grocery store, a quick trip to grab some milk, buying smokes at the gas station, pumping gas in a state where you pump your own gas. (Unfortunately, I doubt gas station attendants are getting sufficient PPE but that's a different discussion.) Things that are too risky: going to church, going to the office, going to a bar, dining in at a restaurant, exercising in the gym. Any sort of closed box with a large number of people for extended periods of time.

If people are able to understand that activities that are too risky should continue to be prohibited until we've advanced our testing capabilities by at least order of magnitude more, and can think of a "quick trip to grab some milk" as a high risk activity, then there is "bubble merging" that can happen that does not result in any change in the growth curve, bit with a reduction in misery. Can we avoid making misery the point of shelter-in-place? Whether people are capable of understanding that is part of that discussion, especially given that "people" includes doctors and people drinking aquarium cleaner.

We still don't want to reopen early, so for the love of god, continue staying away from other people. Heed your local authorities on what is appropriate for your area. Call me a nerd and an armchair epidemiologist (despite not claiming to be an epidemiologist anywhere) because I find the math and the social science that goes into the decision interesting. The alternative is bowing to anti-intellectualism that the current President espouses, where words don't mean things and we're all triggered snowflakes who can't have a discussion. (Not calling anybody out with that, mind you.)
posted by fragmede at 12:34 PM on May 5, 2020 [3 favorites]

Shame people not complying who violate your space. Hope you don't get shot. Don't be breaking the rules when you do it. Push back on those who want to "reopen America". Now is not the time.
posted by Windopaene at 2:28 PM on May 5, 2020

I get very antsy about suggestions like these because it's highly likely that areas with predominantly PoC, Indigenous, and other marginalised communities will be marked as "red" even if they don't actually have higher cases. (This is already happening with police enforcement in Australia and the US - PoC & PoC-heavy neighbourhoods are being targeted with far stronger measures than White people and White-heavy neighbourhoods, even though the COVID19 cases aren't proportionately similar.) It also means that those in the red zones are not going to be able to access necessary help if their main provider just happens to be from a different zone.
posted by divabat at 4:24 PM on May 5, 2020 [2 favorites]

I'm just never ever leaving my house again. That's the only solution.
posted by jenfullmoon at 7:04 PM on May 5, 2020 [1 favorite]

I suddenly remembered the Susan Ertz quote (that I'd somehow thought was from Mark Twain; I had to go track it down):

"Millions long for immortality who don't know what to do with themselves on a rainy Sunday afternoon."
posted by ErisLordFreedom at 10:42 AM on May 6, 2020 [2 favorites]