I could have written a post like this one every other day for the past few weeks. Highlight one of the right-wing blogs I read and talk about their reaction to the Covid-19 pandemic. The story would be the same over and over: a mix of genuine confusion, an even more irrational faith in free market economics than usual and the now standard belief that genuine expertise is the hallmark of deception.
But I’ll highlight the inevitable one: Sarah Hoyt https://accordingtohoyt.com/2020/04/03/assume-a-spherical-cow-of-uniform-density-in-a-frictionless-vaccum/ The truth of the general statement I made above would also be nearly true of Hoyt’s blog. Not quite every other day but nearly so, there has been a post about the virus offering a close to fact-free dissent about the wider view of the pandemic.
The denial isn’t hard to understand. There really is no doubt that measures to reduce social contact reduces the spread of the disease – indeed, that’s almost axiomatic about communicable diseases. There’s also not much doubt that reducing social contact has a negative impact on the economy. Which takes us straight to the dilemma of every nation on Earth currently: saving lives will hurt your economy. A corollary to that is that there really is no immediate free market solution to the pandemic. Give it time and yes, there are fortunes to be made from vaccines and treatments but this current situation is genuinely a big-government kind of problem and hence even conservative governments are trying to buy time with quite severe laws restricting our movement.
For libertarians and pseudo-libertarians this must be nightmarish. OK the actual situation IS nightmarish but for the pseudo-libertarians like Hoyt the world has turned on its head. The route through the next months has narrowed to variations on the same basic policy: massive government efforts to keep the health system running, laws massively restricting human movement, massive government spending (based on borrowing) to stop the economy from collapsing. This is not a war (the pseudo-libertarians quite like war) but it is not unlike a war-footing but without the militarism that the pseudo-libertarians enjoy.
For the piece linked above the frame is a standard denialist line: models are simplifications of complex things and hence don’t capture the complexities and hence must be false and wrong and bad etc etc. Part of that is true. Models are simplifications of complex things and have aspects that are known to be both false and misleading. The simplest example (and analogy – which is cool that an actual example is also a metaphor for itself) is a map. Maps leave out details. A roadmap exaggerates the width of roads for the purpose of visibility. Any model must contain such simplifications and errors because that is the purpose of models.
The situation is even more dire than that though. Not only is every model ever wrong (to some degree) but we have no choice but to use models. Unless you are omniscient being, you can’t know everything. So you HAVE to use models. Your brain uses models, your basic SENSES use less than perfect models that approximate and fill in missing details. It is not unlike the version of the laws of thermodynamics (attributed to either Allen Ginsberg or C.P.Snow – take your pick)
- You can’t win
- You can’t break even
- You can’t leave the game
People get that the first two must be true about any kind of model (cognitive, mathematical, computer-based) i.e. that the model is a simplification and that there will be aspects of the model that are misleading. People don’t always get the last one: you can’t escape models. Which takes me back to Hoyt:
“This came to mind about a week ago as I was stomping around the house saying that anyone who relied on computer models for anything should be shot. My husband was duly alarmed, because as he pointed out, he has designed computer models. At which point I told him that’s okay because his models do not involve people. Which is part of it. Throw one person into a model, and you’ll wish the person were a spherical cow of uniform density in friction-less vacuum.”
The question Hoyt raises unintentionally is if people are not to rely on computer models then what SHOULD they rely on? What is the alternative? Because not relying on models at all is an impossibility. The virtue of a formal model is that they are examinable. Hoyt uses the old joke about the mathematician given the task of helping a farmer but the joke itself reveals a strength of a mathematical model as the butt of the joke. The simplification and hence the way the model departs from reality is overtly stated. The alternative is situations were we use models without realising we are doing so an without understanding how the cognitive model we are using departs sharply from reality.
Luckily for me (if not for the health and safety of her readers) Hoyt provides a perfect example of exactly that kind of unexamined model:
“It’s hard to deny the disease presents in weird clusters. I have a friend whose Georgia County is about the same level of bad as Italy. Which makes no sense whatsoever, as they have no high Chinese population. And while the cases might be guess work (with tests only accurate AT MOST 70% of the time, it’s guesswork all the way down) the deaths aren’t. The community is small enough they all know each other. And they’re losing relatively young (still working) and relatively healthy (no known big issues) people.”
Hoyt is still stuck with a mental model of Covid-19 as a “Chinese” disease — as if somehow the novel coronavirus has a memory of where it first infected humans. Spread of the disease has long since moved well beyond travellers from China. For example, I believe in Australia more cases originated directly via travellers from the USA than from China. Mind you, remember this a person who puts every effort into refusing to believe that there can be such a thing as unconscious biases (at least among people she approves of).
Having robustly asserted how people aren’t spherical cows, Hoyt then promptly spends multiple paragraphers generalising about New Yorkers and Italians and so on. More flawed models.
That takes us to Colorado. Colorado, Hoyt assures us, is different. Now that is clearly true. Colorado is not Italy and it is not New York and some of those differences do matter for the spread of the disease. It is a less densely populated state without a doubt. Hoyt argues that because Colorado is different then the rules should be different.
“So, why are the same rules being applied to both places? AND why are both places treated exactly alike? And why are both places assumed to be on the same curve as Italy or Spain or Wuhan, places and cultures, and ways of living that have absolutely nothing to do with how we live or who we are? And here’s the kicker: if you allow states like Colorado and others that naturally self-distance to go about their lawful business, not only time but more money will be available to study the problem clusters.”
Here is the real kicker. Models are imperfect (by definition) and those imperfection can be misleading (by their very nature) and you can’t NOT use models of some kind or another BUT we have a way of minimising the mistakes we make. The method is simple but it has taken us millennia to work it out: we check the outcomes of our models against data and observation. Now even with data we still have models (sorry, they are inescapable) but we have ways of checking our conclusions against others.
Colorado isn’t a mysterious far away planet. We can literally go and see how Covid-19 is progressing in the state. I’ll use the John Hopkins University visualisation tool for tracking confirmed Covid-19 cases that is available here: https://www.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6 The tool allows you to drill down to state (and within state) data in the USA.
Colorado (pop. 5.696 million) currently (April 4 6:50 Sydney time) has 3,742 confirmed cases of Covid-19. For comparison, New South Wales (pop. 7.544 million) has 2,389 confirmed cases and that’s with long established Chinese communities (that Hoyt seems to regard as the only risk factor) as well as Sydney being a major cruise ship destination (an actually pertinent risk factor). Colorado does have major ski resorts* and I suspect we’ll get a better sense of the role they played in the pandemic in the future.
Yes but…as I said, even data relies on models of one kind or another and maybe Australia and Colorado are using vastly different diagnostic criteria or maybe it is due to vastly different testing regimes. I might genuinely be comparing apples and oranges. Sadly, we can reduce (but not remove) disparities in reporting by looking at a more sobering statistic: deaths.
According to the John Hopkins University dashboard New South Wales has 12 confirmed deaths. That’s a tragic and worrying amount. Yes, many more people die from all sorts of other causes but these deaths add to that total or mortality and the progress of this pandemic is far from over. That’s just the beginning of the numbers.
Let’s compare with Colorado (there is also state specific data here also https://covid19.colorado.gov/case-data). From the same data source Colorado has had 97 deaths so far. It’s when I saw that number that I shuddered and decided that I’d write this post rather than just shake my head at Hoyt’s nonsense. I knew things were bad in some parts of the US but I’d assumed that some of the denial I was reading was because the writers of this toxic nonsense were in states were the wave of the pandemic was still to hit. Ninety-seven deaths, shit. I keep looking at that number and knowing that there other places in the US where the numbers of deaths are being under reported particularly for vulnerable communities and shuddering at what might be the true scale of thins.
Now sure, maybe the differences in testing and diagnostic criteria and data collection are so different between NSW and Colorado that the number of cases is incomparable BUT they would have to be significantly different in two different directions simultaneously. That is, if NSW are under-reporting the number of cases compared to Colorado then the case-fatality rate in Colorado is even worse when compared with NSW. I’m not making the comparison to say which state is somehow doing ‘better’ (it’s not a race or a competition) but simply trying to get a sense of what I can see HERE and compare it with where Sarah Hoyt is. It is undoubtedly a crisis here and we’ve got a conservative government in power at the state level and the national level and heck, both of them if they had an excuse to cut spending and pull back on entitlements and let business run wild they would and you know what, they aren’t and in fact they are doing the opposite. That’s not because they have had a sudden ideological conversion to policies they have derided for years but because massive government spending is the ONLY way to keep the economy going. When conservative ideologues rush to implement free government funded childcare it is safe to assume that they felt they had no other choice.
The morbid irony here is that Hoyt is ignoring her own advice. Rather than just look at Colorado and consider whether that state, regardless of what is going on anywhere else, is in the midst of viral outbreak and in grave danger and what action in such a circumstance the state government should take (hint: major restriction on movement and social contact to keep hospitals going and to give time for treatments and vaccines to be developed) she is insisting that because Colorado is not New York it can’t need the same measures as New York. It’s a compounded level of illogic.
Strip everything away from that piece by Sarah Hoyt and what you are left with is the common theme that captures so much of the train of political thought that joins Ayn Rand to Trump to Jordan Peterson: the desire to dress up wishful thinking as something other than a demand that reality should accord with their personal desires.
There’s no conclusion. Stay safe. Wash your hands. Think of others. Be kind. Don’t spread nonsense.
*[To be fair New South Wales does have ski resorts as well but during the start of the pandemic it was 1. summer here and 2. they were on fire.]
