Image: Angelo Pantazis/Unsplash

Why our roads got proportionally more dangerous during the COVID pandemic

Road fatalities dropped, but not as much as they should have. And as for cyclists? The numbers aren't great.

Don’t miss out on the latest CyclingTips updates.

Jump To Comments

The COVID pandemic reshaped our lives in myriad ways. From how we worked, to how we interacted with one another, to how we lived our lives more generally. And while COVID is still very much part of our lives, we’re now getting a clear idea of the significant changes the pandemic has wrought.

Among the wealth of COVID-related research coming out around the world is a body of work on the way road safety changed during the pandemic – a subject close to heart for all of us that ride bikes.

Among the work in this space is a new discussion paper from Australia’s National Road Safety Partnership Program (NRSPP), a program “about saving lives without the red tape”, based at the Monash University Accident Research Centre in Melbourne. The NRSPP report considers road fatality data from the period of March to July 2020 – a time in which much of Australia was under various forms of lockdown – and pulls in a bunch of research findings to make sense of how lockdowns affected road safety.

As the report’s author Dr Sarah Jones writes, the “first immediately identifiable impact of the pandemic is a reduction in road fatalities,” which were down to their lowest levels since 2012. This is to be expected: with so much of the population locked down, fewer kilometres were travelled on our roads, reducing risk exposure and, ultimately, the road toll. For the same reason, the number of on-road injuries also fell dramatically.

Worryingly, though, even with the country on lockdown, Australia wasn’t able to meet a National Road Safety Strategy target of a 30% reduction in road fatalities, throwing into question how realistic those reduced targets even are.

Note the black line (total road deaths) didn’t even drop to target levels (dotted cyan line) during lockdown.

But the biggest takeaway from the data is that the reduction in road fatalities is not commensurate with the reduction in kilometres travelled.

This finding comes about by using fuel sales as a proxy for total kilometres travelled. The assumption goes that the less fuel was purchased, the fewer kilometres that were travelled. While petrol sales in Australia dropped by 21.4% from March to July 2020 (compared to 2017), the road toll only dropped by 10%.

The implication here is troubling: while there was a reduction in road deaths, there should have been a much greater reduction given how much less driving was going on. Only in New South Wales and the ACT did the drop in petrol sales and road fatalities line up, while in Queensland, there was a 17.4% reduction in petrol sales but an 11% increase in road fatalities.

So what’s going on here? Why did the roads become proportionally more dangerous during lockdown? According to the NRSPP, there are four main factors that could be responsible.

1. Greater inattention / distraction on the road

As Jones writes, the COVID lockdowns of 2020 were characterised by fear and uncertainty. When would vaccines be available? How long would we be locked down for?  Would life ever go back to normal? It’s no surprise that, as the Australian Institute of Health and Welfare notes, “stress, confusion and anger [were] commonplace as a result of the pandemic.” By Jones’s estimation, “these emotions can create cognitive deficits that reduce a driver’s mental ‘bandwidth’ for attention on the road,” which, in turn, can lead to greater frequency of on-road incidents.

That might seem like a loose explanation, but it’s worth noting previous research in this space. A driver who’s been through a recent divorce or separation has a four-fold risk of a serious traffic crash that’s at least partially their fault. Financial stress has also been linked with a higher-than-normal crash risk.

Not only that, but an Australian Road Safety Foundation survey from April 2020 found that 78% of drivers admitted to breaking a road rule during lockdown, with 39% attributing that to not paying attention and 30% citing a brief lapse in judgement.

2. An increase in the number of vehicles as a workspace

The volume of online purchases significantly increased during the pandemic, which meant more drivers were required to deliver all of those incoming goods, whether they be food and drink, or other packages. In Australia, around 2,000 postal workers were retrained as van drivers to cope with the increased delivery load, and of all the delivery workers engaged in work through an app, 59% joined to support themselves through the pandemic.

Crucially, there are few rules relating to the safety of Uber Eats drivers, say, compared to those who drive heavy freight vehicles. 

“A restaurant looking to compensate for loss of foot-traffic through diversifying into home delivery has no direct obligation to develop a delivery model that limits incentives for drivers to speed and/or drive fatigued,” Jones writes. “Most Uber Eats delivery drivers have jobs elsewhere. Drivers often ‘multi-app’, with no obligation on app operators to ascertain how much work a driver has done elsewhere or how fatigued they may be.”

That may sound like a long bow as well, but consider a 2019 multi-government report entitled “Vehicles as a Workplace – Work Health and Safety Guide”, which notes that “vehicle use is by far the most significant  contributor to work-related traumatic injury.” In fact, 64% of worker traumatic injury fatalities since 2003 involved a vehicle.

In Jones’s eyes, all of this “may somewhat explain why road fatalities did not fall at the rate we might otherwise have expected: the road became a ‘workplace’ to hundreds of people not necessarily trained, experienced and proficient in managing the occupational hazards of that workplace.”

3. An increase in vulnerable road users

There’s no doubt the number of people riding bikes during the pandemic increased. You only had to head out to your local park or bike track to see everyone on bikes they’d just bought or dug out from the back of their shed. In many cases, money that could have been spent on travel or going to events was used to buy a bike, leading to a new record for bike sales in Australia: 1.75 million imported in 2020-21 vs the previous record of 1.42 million in 2007.

While the growth of the cycling cohort was heartening to see, there was a grim reality to the increase in rider numbers: the National Office of Road Safety reports that cyclists were the only vulnerable road user group to experience an increase in fatalities in 2020 vs the 2008-2010 baseline. The same was true of serious injuries, vs a 2018 baseline (it’s not clear why those baseline years were taken and why they were different to one another).

There was a 44% increase in the number of rider fatalities in multi-vehicle crashes (i.e. the cyclist and another vehicle), while the number of cyclist fatalities in single-vehicle crashes (i.e. just the cyclist) decreased slightly.

The cyclist fatality increases were limited to Queensland and Victoria which, particularly in the case of Queensland, might somewhat explain the significant difference between fuel sales (as a proxy for distance driven) vs on-road fatalities in that state.

So what’s the reason for the overall increase in rider deaths? A big part of it may just be down to the increase in cyclist numbers. It could also be that “the proficiency of new or resuming cyclists was low and this accounts for the high fatality and injury count,” Jones writes. “The non-involvement of buses and trucks, which are generally over-represented in cycling fatalities, tends to support this theory.”

Perhaps the factors already mentioned above – reduced attention from drivers and an increase in vehicles as a workspace – also contributed to more dangerous conditions for cyclists. And then there’s the fourth and final possible explanation discussed by Jones …

4. Proportionally more risk-takers on the road.

“During the pandemic, road users may have reasoned that on-road enforcement would reduce as law enforcement resources were redeployed into other tasks or kept from situations of potentially high transmission,” Jones writes. “In at least one jurisdiction [Queensland] regulators were publicly candid that this was, indeed, the case.” 

Sure enough, random breath tests (RBTs) in Australia dropped from 15 million in 2019 to less than 7.5 million in 2020. Meanwhile the percentage of positive RBTs doubled, from 0.4% to 0.8%.

“This suggests drivers were taking more risks on the road during the pandemic,” Jones writes. “It may be that the prevalent anxiety and uncertainty discussed earlier in this paper prompted greater alcohol consumption, although the data on this is not definitive.”

It’s not just more drunk drivers either. An Australian Road Safety Foundation survey revealed that one quarter of responding drivers self-reported taking greater risks on the road during the pandemic, with 17% admitting to speeding more.

Plus, it seems reasonable to assume that those who were out driving when they shouldn’t have been are the sort of people less inclined to follow rules in general. Including the laws designed to keep road users safe. “This cohort would likely be less influenced by stay-at-home orders and more likely to flout rules relating to curfews, geographical restrictions and household ‘bubbles’,” Jones writes. “Being less rule-governed than the general population, they are also more likely to speed and take other risks on the road.”


So where does all of that leave us? To summarise, road fatalities dropped during early 2020 as most people followed stay-at-home orders. But the fatality rate didn’t drop in proportion with the reduction in the amount of driving done, which is concerning. We should have expected a further reduction in the number of road fatalities.

Jones’s discussion paper offers potential explanations for this discrepancy, but question marks certainly still remain. This isn’t a peer-reviewed research article but a discussion paper that leans on assumptions to make sense of the data. And while those explanations make some intuitive sense, the reality is we might never get the full picture of why things changed on the road as a result of COVID. 

From a cycling point of view it would be good to understand more about why the number of cyclist fatalities increased during lockdown. Was it simply down to having more riders on the road, or was there more to the story?

Ultimately this report was written with the goal of “informing policy responses to the next crisis, be it pandemic, fire, flood or the effects of climate change.” Hopefully the findings here can go some way to ensuring better outcomes when the next life-changing crisis engulfs us, whatever that might be.

Editors' Picks