The HSE have released their Workplace fatal injuries in Great Britain, 2021 report with provisional data showing 142 workers killed in workplace accidents in 2020/21. Doubtless you’ve already seen the data.
Although only provisional data at this point, revisions at this time next year are likely to be minimal and 2020/21 will see annual fatalities confirmed somewhere in the 140-150 range. The same as every other year in the past decade – except 2019/20 which has now been confirmed as having 113 fatalities.
When the provisional data for 2019/20 was released we did ponder whether this was the year that finally saw fatality rates fall significantly below what had become a plateau over recent years. We were optimistic that those results were statistically significant and not an anomaly brought on by the onset of the pandemic.
To see the latest provisional numbers of fatalities at 142 suggests that our optimism was misplaced. Or was it?
How do we compare previous years with the last two, both of which were impacted by lockdowns? The HSE refer to the ONS Labour Market Statistics (LMS) which show total actual weekly hours worked (millions) as:
- 2018 = 1037.6
- 2019 = 1052.1
- 2020 = 943.7
These are for the calendar year whereas the HSE injury reports are presented over the tax year (Apr to Mar) where the corresponding number of fatalities was:
- 2018/19 = 149
- 2019/20 = 113
- 2020/21 = 142 (provisional)
The quarterly LMS data allows us to see that the weekly hours worked in 2019/20 is slightly higher than in 2018/19 and that 2020/21 is just over 10% down on 2019/20 – much as in the data above. In which case, 2019/20 remains an anomaly and 2020/21 looks as if the fatality numbers are slightly higher than they should be (given 10% less work) although this is unlikely to be statistically significant.
So, working hours doesn’t explain the anomaly that is the lower fatality rates in 2019/20. Given the 10% reduction in work carried out in the heavily lockdown influenced 2020/21, how can we explain the return to normal in the number of fatalities? That one, at least, seems pretty straight forward in that furlough mainly affected ‘office workers’ least at risk from workplace injury.
The anomaly is 2019/20 and, according to the HSE, the fall in the number of fatalities is statistically significant when compared with both the prior year and the average of the previous five years. If not worked hours, then what?
Let’s look at how 2019/20 compared with the most recent year and the average of the previous four years by sector.
2019/20 is the central bar so we’re looking for sectors where that bar is significantly below what came before and afterwards. Agriculture fits and has the largest numbers followed by Manufacturing and then Other but nothing leaps out as being the single cause of the anomaly.
What if we look at fatalities by age?
A similar scenario with age groups 45-54 and 65+ standing out but little reason to suggest any causal relationship between a single age group and lower fatalities in 2019/20. The reduction in age 65+ fatalities is likely to be connected to Agriculture based on previous analysis but little more can be derived for these purposes.
Do the causes of fatality shed further light?
Possibly. The Other category almost entirely accounts for the anomaly that is the 2019/20 reduction in fatalities. What causes constitute ‘Other’?
A myriad causes, but still nothing that can genuinely explain why 2019/20 has a statistically significant lower number of workplace fatalities than prior or, indeed, subsequent years.
Remembering that 2019/20 was essentially pre-lockdown in its entirety, here are some hypotheses to explain the anomaly:
- the number of fatalities in 2019/20 is simply an outlier on the bell curve
- there was a, yet to be uncovered, reporting issue
- 2019/20 did represent a significant breakthrough in reducing workplace fatalities. What if, in returning to previous fatality rates, it is 2020/21 that is the anomaly?
Is it possible the last of these could be the case? Did workers’ individual risk assessments change during 2020/21 and did that lead to changing workplace behaviour with negative outcomes?
The lens of history will, no doubt, shed further light. In the meantime, what do you think? Please leave a comment below.If You Like This Post, Please Share It!