The insanity continues in supermarkets in Australia with ongoing cases of verbal and physical abuse of supermarket staff. Reports such as this one also include photographs of massive queues at supermarkets across Australia. Personally, I haven’t experienced this phenomenon. True, with the exception of today, which I shall explain in a moment, I have only shopped at one supermarket, but there have been no queues at that supermarket on the multiple times that I have visited the store. Even more surprisingly, parking spaces have always been readily available.
Today, I had to venture further afield to do my shopping because I needed to visit a pet store near to a different supermarket in order to pick up my online order of cat litter. I know, cat litter concerns in the midst of the Covid-19 virus but the fact is that my local pet store had run out of the brand that I usually purchase. Anyhow, I wondered if I would come across car parks filled to overflowing along with huge queues outside of the supermarket. Such was not the case. The car park was little more than one quarter full and there were no queues outside the supermarket.
Whilst I did not experience the mass shopping reported in the news stories, I did have my first experience of another popular news topic, an abusive and aggressive shopper. This is the first time that I have experienced this phenomenon. I had to get some change to unlock a shopping trolley and so I went and queued up at a cashier line. The woman in front of me was overly large and exuding anger from every pore in her body. She was utterly and irrationally angry because the store had placed a restriction of purchasing no more than two of the products that she wanted to purchase. Said large woman wanted more than two of the products and she repeatedly complained that there should have been a sign telling her what she was allowed to buy.
A male staff member did absolutely the right thing by letting the customer know that he would personally ensure that every aisle had the relevant information about the purchasing limits for the various products in the aisle. The woman huffed and puffed and argued again that signs should have been put up. She had already made the same point at least ten times and it would be fair to say that she was being somewhat repetitive. A female shopper standing serenely nearby tried to diffuse the situation by smiling graciously at the angry woman and offering that everyone was having difficulties. I was about to look for a security guard when the angry woman spat out, “I know, well at least I managed to get a few things.” She then stormed out of the supermarket much to everyone’s relief.
I did my shopping and, guess what, the shelves were lined with just about every product that one could wish to purchase with the exception of toilet paper of course. So, goodness knows what the angry woman had to be angry about. When I came to pay, I chatted with the checkout girl, asking her how her day was going and sympathizing with her regarding the difficult conditions in which she was working. As she came to ring up my two packets of sponges – a contingency measure given that I cannot get any toilet roll – she told me that I’d already bought a pack of cleaning cloths and had reached the limit on that line of products. I said, “No worries, I’ll leave them” and then she took a quick look around and said, “Pay for all this and then I’ll put these through separately.” And so she did.
Being in Australia, which still has relatively few Covid-19 cases, I am obviously interested in what the future might hold with respect to contagion rates and fatality rates. There’s not going to be anything particular scientific about this analysis. Rather I am just going through an exercise of looking at data because I want to try to gain some idea of what it looks like to do data modelling. More specifically, I want to try to get a feel for how many variables one would have to take into account in making projections about the likely spread of the virus in Australia. This point interests me because I have so many different predictions.
There is a massive difference in Covid-19 deaths as between Italy and South Korea and I thought I’d take a look at why this might be the case. First, South Korea has carried out far more Covid-19 tests than Italy, a fact which may help to identify and isolate cases so that the individuals with the virus do not spread the virus to other people. So, there’s one variable that will impact on the likely spread of the virus in Australia. Number of tests carried out. If you take a look at the testing rate data on the government website then, it would seem to me, that testing rates are incredibly low. For example, New South Wales, the State with the highest testing rate as carried out around 52,000 tests with a population of 8,129,000 people. The Australian Capital Territory has carried out around 2000 tests with a population of 427,4000 people.
Italy has a much higher incident rate in persons over 60 years of age as compared with South Korea where the outbreak has occurred amongst much younger people. In South Korea women account for a greater percentage of those infected than is the case in Italy so South Korea is seeing more cases in younger females. If we add smoking into the mix then smoking rates in Italy and South Korea are about the same. However, far fewer women smoke in South Korea as compared with Italy. So, in summary, “South Korea has an outbreak amongst youngish, non-smoking women whereas Italy’s disease is occurring among the old and very old, many of whom are smokers.” So, where we have some more variables, age, sex and smoking rates.
If we’re trying to think about what might happen in the future with respect to the spread of Covid-19 in Australia and the number of fatalities then we could consider a range of variables drawn from evidence about the variables that have impacted Italy and South Korea. We would then look at demographics from Australia. This would be a very simplified line of argument but I am trying to keep matters simple because my aim is just to explore a few variables in order to better understand how data modelling might work. Based on the data from Italy and South Korea we would be broadly interested in age data, relative proportions of male to females in the population and data around smoking. This data is readily available.
According to the 2016 census, Australia’s population was 23,401,892. 49.3% of the population was male and 50.7% of the population was female. The median age was 38. 7,953,6583 people or 33.98% of the population were aged 50 years or over. 947,931 people or 4.05% of the population were aged 80 or over. In 2017-18, just under one in seven (13.8%) or 2.6 million adults were daily smokers, while a further 1.4% of people also reported smoking. 18.4% of men smoked and 10.5% of women smoked in 2017-18. The smoking rate amongst men declined age 55-64 years before eventually dropping to 5.1% at age 75 years and over. For women, one in ten (10.4%) 18-24 year olds smoked daily the smoking rate dropping to 7.5% for 65-74 year olds and 3.7% for women 75 years and over.
We also have more nuanced data telling us that, “People aged 20 to 44 years made up 38% of the combined capital city population, compared with 30% of the population in the rest of Australia. This reflects the attraction of younger adults to education, employment and other opportunities in capital cities.” Furthermore “Older adults aged 45 years and over made up a smaller proportion of the population in capital cities (37%) than in the rest of Australia (45%).” With this data we might begin to conjecture about the likely spread of the virus in capital cities noting a higher percentage of younger people in the cities as compared with the country where there is a higher proportion of older adults.
To give another example, in “both capital cities and the rest of Australia, there were higher proportions of females than males in older age groups. The difference was most marked among the population aged 85 years and over, and is due to the longer life expectancy of female Australians.” So, if we were looking for another possible pattern we might consider the susceptibility of women aged over 85 in cities and in the rest of Australia. As a final example, there is data on the regions in Australia that have the highest median ages, these regions being popular retirement destinations. The Hunter valley, for example, has a median age of 62.7 years. So again, one might look to retirement destinations as like areas for Covid-19 infections.
Basically I’ve gone through this process of looking at data and variables because I wanted to understand the complexities of trying to model how the pandemic might unfold in Australia. Beyond the data, there are also variables that either will never be fully known or will not be known at all. Perhaps most importantly in terms of trying to predict the spread, Australia has no real idea how many people are actually infected with the virus and so a baseline cannot be established. Additionally it is impossible to know how many people are conforming to the lock down laws. I doubt that we will have any idea about the percentage of the population that scrupulously practiced social distancing. The same will hold true for people practising good personal hygiene habits. So, overall, I’m wondering if data modelling amounts to much more than guess work.
The other way to look at the future is not in terms of interrogating the data for one’s own particular country but in terms of what is likely to occur in terms of the world as a whole. There are best and worst scenario predictions in this respect, albeit written from an American perspective. The worst case scenario is that the virus remains rampant, millions die and the world economy plunges into a huge depression with millions of people becoming unemployed across the globe. The best case scenario is that the virus is no longer rampant, deaths number in the thousands, are limited to particular age groups and those with pre-existing medical conditions – the current scenario – and the economy rebounds.
In terms of the best case scenario, I have written in other posts on Severe Acute Respiratory Syndrome (SARS) and Middle Eastern Respiratory Syndrome (MERS). Cases of SARS and MERS did not even approach the number of cases that we have seen with Covid-19 and both SARS and MERS have “petered out.” It seems unlikely that this will be the case with Covid-19 given that the virus has already taken hold around the world and that the virus is considerably more contagious than previous viruses. It would, therefore, seem to be more sensible to think in terms of what can be done to control or contain the spread of the virus. Governments obviously have a crucial role to play in control and containment but individuals can also play a significant part in terms of the variables that I have identified in this post.
First Published March 21st, 2020