Demand shocks and toilet paper
One of the most interesting things to happen in the COVID-19 pandemic was a run on toilet paper in Australia. This shortage is already past us now and households have far more toilet paper stocked up than usual. I think in a while this event and the causes of it will mostly be forgotten, but there's some interesting things about this event that hopefully aren't lost to history.
Toilet usage is surprisingly culturally specific with many different cultures having different habits, there's even a world toilet day to bring awareness to the major sanitation issues that still exist world wide. One thing I can say for sure is that Australian culture is very much a toilet paper culture and in early 2020 this culture faced a surprise shock.
So what happened?
At the beginning of the pandemic household toilet paper supplies at retail locations started to fall (more about why a little later). This triggered a run on the remaining retail supplies that then further deepened the hoarding behavior. This feedback loop was noticeable. What started as a supply shock turned into a full on panic and people started hoarding any toilet paper they could find. I remember before the panic went completely mainstream I got some toilet paper to have on hand just to give to family and neighbors in the case that they ran out, effectively I was getting more than usual at this point mostly just due to a reaction to the panic buying. This collective behavior led to empty toilet paper shelves at stores which fed back into the panic buying feedback loop.
Which then led to people using social media to tell other people about how shelves were empty and increasing the panic. When I was discussing this situation with Itsi he astutely pointed out that since toilet paper is bulky any change in stock levels will look very obvious and this makes for easy photos to be shared when these large shelves are empty, more about the economic impact of the bulkiness of toilet paper items later.
As is the case in many situations where there's huge volatility in pricing you get people starting to do things just to quickly cash in. In this case some people started to hoard toilet paper just in the hopes of scalping it later. I remember a clear "peak toilet paper" moment when I was in a local store that just got a palette of toilet paper delivered, they didn't even bother to shelve it and just left it in the middle of the floor, customer limits only came in place the week after. As I was leaving the store I remember clearly hearing someone on the phone talking about this find to a friend "I know this is on the other side of town but they just got a delivery and you should get here quick, there's still boxes of toilet paper here". You could tell that people were willing to put up with substantial inconvenience to get toilet paper and things were starting to get very silly:
This was a fantastic example of a demand shock. I find the various details of this most interesting, especially the not so obvious ones.
Why did the panic impact toilet paper first?
There's a ton of stories about the psychology of the panic buying, but far fewer about what caused the panic in the first place. Also why was it toilet paper that came to run out first? At the beginning when there was significant doubts about the pandemic situation people weren't sure how they would be able to get supplies, or even if they would be able to get supplies in the case of a severe lockdown. Against this backdrop of uncertainty there was a sharp increase in demand for any items that would be needed in the case of surviving a lockdown. Demand for things like non-perishable food went up a lot. And of course demand for toilet paper as we know went up.
One of the most interesting things about this event is that Australia is capable of manufacturing more than all the toilet paper it consumes locally. Sufficient manufacturing capability is the usual state of affairs, but yet in the height of the panic people didn't realize this was the case. Pushed forward by news cycles that promote clickbait and outrage, news of the shortages spread extremely fast and amplified the panic, people really did think they would run out of toilet paper. Reporting that would encourage more sensible behavior was far harder to come across. Eventually people came to understand that production wasn't actually an issue and the panic started to subside somewhat. It took a bit longer for shelves to get more stock, but more about why that happened later.
At first sight it seems like this is just a demand shock caused by panic buying, and while herd psychology was a big factor it certainly doesn't tell the whole story. Availability of some items fared far better than others, and that was because there was a really important supply shock component to this panic for some items on top of the demand shocks. A particularly interesting thing is that there was likely no significant change in overall bathroom usage for the average person before and after the pandemic started, the main difference being where people were going to the bathroom. The toilet paper used at homes is quite different to that used in commercial places as you can see in this image:
The production lines for these different toilet paper products are not exactly the same, and as such there's a significant switching time (if it is even possible at the given factory) for a factory to make less of the commercial paper and more of the home paper.
Add to this that a large amount of the office supplies were left sitting at offices that were closed due to lockdowns and this reduced the supplies even further. Even if you didn't have the right dispenser you couldn't easily get the toilet paper from your workplace if your workplace was closed by the lockdown.
One thing I find particularly interesting is that even when the panic started getting more pronounced the supermarkets didn't tend to stock any of the commercial toilet paper. This simply didn't seem to be an inventory item at many stores, which makes sense when you consider that these are bulky items and keeping them in stock would take a lot of storage space. I suspect keeping the inventory low on bulky items, or just not stocking them at all, is a consequence of reduced storage space and Just In Time logistics strategies, something I will cover later.
Even at the peak of the panic buying for toilet paper there was still availability of commercial rolls. I have one commercial roll at home (the one in the picture) because at the time it was almost impossible to find regular rolls in stock. This state of affairs didn't last all that long and the non-commercial toilet paper supply grew to meet this new work-from-home "normal".
The perils of "just in time" storage and warehousing
From what I understand similar toilet paper shortages occurred in other countries that have a toilet paper culture, in the places where the supply ran out I suspect the overall approach to supply chains had a lot to do with it. One of the things that had really increased the panic was how quickly the original stock at stores ran out combined with the difficulties in restocking.
High real estate prices have greatly pushed the prevalence of Just In Time deliveries in areas where storage space would be expensive. The approach to trying to deliver things immediately before consumption is often known in the industry by the acronym JIT and isn't particularly new. People have been discussing "push" vs "pull" strategies for logistics and supply chain management for quite some time now.
If you are opening a store in an area with high real estate prices per square of floor space you have a strong financial incentive to reduce the amount of warehousing space. Reducing the amount of storage space at the stores comes with a substantial reduction in overall system resiliency and puts more stress on the supply chain if it is to feasible at all 1. One of the reasons toilet paper ran out so quickly at the stores is because of it's physical size. Compared to other inventory items toilet paper is very bulky and in "normal" times stocking a lot of it would give you very poor monetary utilization of the warehousing space you were using. Many stores would have a large percentage of their stocks of toilet paper sitting on the actual retail space floors for this reason. Deliveries would then be calibrated to try to restock the shelves at the same rate of average consumption.
One particular aspect that has become much more popular in recent times is statistical demand forecasting. Effectively what this does is use past data to inform decisions about warehousing and supply chains. These techniques have found use in many areas from determining stock levels to which warehouses you store the stock in all the way to the physical layouts of the warehouses storing the goods. This forecasting has become more of a mainstay because of the nature of preparing inventory specifically for cyclical events, for example in Australia many stores prepare for events like Christmas and Easter months ahead of time to get various seasonal stock items in place. This likely makes those calendar events much more durable but what risk does this generate to other scenarios? In some senses due to the fragility they they impose upon themselves with some of the "optimizations" done to the supply chain they have to plan this far out in advance to make sure there's a seamless transition between two very different inventory setups. I'm not sure how many of these systems are built with resiliency as a goal, some I've seen definitely reduce it in teh face of intermittent events.
The major supermarkets in Australia, and elsewhere, have tended to do things that will increase the efficiency of their supply chains in "normal" times that may have reduced the resiliency of those systems to shock events like this.
I believe this supply shock has been an extremely good reminder of how modern supply chain practices can amplify shortages. Luckily this time around it mostly impacted availability levels of items like toilet paper. Toilet paper had no long term supply issues and the very temporary shortage didn't cause any lasting harm. Hopefully people will learn from this and improve the resilience of these systems so that another shock like this doesn't happen with a more critical item.
If you are interested in getting help making your supply chains more efficient in a robust way by applying mathematical supply chain optimization feel in contact by emailing me (janis at scrygroup.com).
Panics and demand inelasticity
Commercial toilet paper rolls are made with the specific goal of reducing the overall amount of time needed to replace the rolls by putting far more paper on each roll. This means that the dispensers for this paper are also of a different size. The economics of commercial toilet paper rolls is driven by trying to reduce the amount of wages spent on changing them. Something I've noticed travelling abroad is that places with higher incomes tend to have higher tech bathrooms, maybe some places have similar toilet paper supplies across all locations?
In any market economy you'll have situations where a fundamental change will the equilibrium of prices and then put pressure on the supply and demand. In this particular case the demand for commercial toilet paper plummeted while home toilet paper went up. When there were signs that the market couldn't flex to seamlessly meet this change in demand the panic started. Apparently some of the toilet paper plants were already "optimized" to be running 24/7 hours already, which meant that scaling up operations wasn't as easy as people may have thought. One thing that's interesting in this case is that there was a significant inelasticity due to logistics issues when the demand suddenly changed at a national level. This was greatly exacerbated by the trend to have more efficient but less resilient supply chains.
Sure people could hoard toilet paper to then attempt to sell it back to at higher prices, the problem with this is that it doesn't increase the real supply and tends to exacerbate the state of panic by reducing availability further. Because this sort of behavior destroys utility for the country as a whole it is mostly frowned upon and as such many laws exist to try to make these behaviors illegal. For a situation with a very temporary inelasticity of supply, hoarding behavior doesn't allow price signals to increase production but rather just temporarily inflates prices. If the item is an essential item then the utility gained by the hoarders is much smaller than the utility destroyed for the people who need the essential items but now do not have availability. It is an attempt to prevent the situation where utility is destroyed by hoarding that has caused anti-profiteering laws to exist.
These laws probably do prevent some people from engaging in profiteering behaviors (who otherwise would) so they are likely a win when a supply shock is very temporary. However these laws do remove some of the signalling that increasing prices can provide to stimulate supply. So if a supply shock is part of a longer term event the anti-profiteering laws could have the unintended consequence of reducing the profitability of those who legitimately would like to raise the supply. In other areas impacted by the crisis this has turned out to be a far bigger issue.
There's been many supply shocks across history, I wrote this because I thought that in the future it might be interesting for people to read a contemporary account of the events of a particular supply and demand shock.
Already in May 2020 the shelves are mostly restocked with toilet paper in supermarkets. Panic has subsided and the great workers of Australia have done a great job in increasing the supply of home toilet paper such that everyone has enough. Overall I think Australia has done well so far in handing this crisis, if history shows that the worst of it was the brief interruption to toilet paper supplies then the country would have done particularly well.
Given the over-reliance on oil for powering modern transport I'm genuinely terrified by the rise in JIT logistics for critical supplies like food. This is because a this turns a single oil supply shock from an economic annoyance to something that could have catastrophic consequences for well being. In particular Australia has been carrying insufficient reserves of oil for me to feel comfortable with an approach JIT logistics that are only feasible with an ample, and reliable, supply of oil. Australia currently has only one month of reserves and very little in the way of proven local oil reserves. In normal times the "efficiency" of oil based transport results in extracting value from increasing risk by not having to service the costs of maintaining other transport and logistic options or maintain a larger position in oil storage. But the downside risk is very severe and I'm surprised that nobody seems to talk about this precarious supply situation. It seemed to take the greed of cashing in on the oil super contango of 2020 to get anyone to want to increase the level of national reserves. For JIT logistics not to be a terrible gamble on the availability of oil, diversification of transport energy sources or increased investment in energy sources that could be burned in internal combustion engines seems sensible. This seems to be a situation where introducing some capacity for transport that could be powered by say electricity would greatly improve the overall resiliency of the food distribution system. ↩