Proving Safety: Observations on “Worker Insights”

By Greg Smith, author of “Papersafe” – first published HERE

Recently, there has been some discussion around the use of “worker insights” as a measure of safety performance, and even as a measure of legal compliance and demonstration of due diligence for the purposes of Work Health and Safety (WHS) legislation. In part, this seems to come from processes developed through a methodology or tool described as the “Due Diligence Index” (DDI).

The DDI, rightly in my opinion, criticises traditional safety measurement based on lag (injury rate) data. However, it also makes several claims in respect to safety assurance and due diligence, which do not appear justified, and the proposed solutions to injury rate data are open to the same corruptions as injury rate data itself.

I want to explore these thoughts through the lens of worker insights as described in the DDI, but before I go too far ahead, I want to give one example of the concerns that are likely to arise from uncritical application of the DDI.

On page 6 of the DDI, one of the criticisms made of injury rate data is that it is open to manipulation (Standard & Guidance for Use Due Diligence Index Council, Version 3.0 Dec 2021).

I do not disagree with this, but there is nothing in the DDI, which is not, of itself, open to manipulation. Indeed, when we look at worker insights as an example, given the level of touch points and potential layers of filtering, if anything worker insights are more prone to manipulation than injury rate data.

Worker insights – the process

As I read the DDI process, the mechanics of worker insights require the collection and analysis of the:

• number of hours worked;
• number of worker insights performed;
• number of worker insights effectively closed out;
• worker insights experience evaluation feedback for workers and leaders; and
• worker insights experience feedback rating weighting.

The mechanics of this process does raise several questions. For example, does the close-out of a worker insight mean the close-out of all the action items which might arise from the worker insight process, or a close-out of the “insight collection” process? Obviously, these are two very different things.

An organisation can speak to its workers and get information from them, but then do nothing about it. Does this qualify as a closed-out worker insight?

And what does “effectively” mean in this context – and who gets to make the decision?

If an organisation can manipulate its injury rate data, I do not think there would be too many impediments to manipulating an “effective” close-out of a worker insight process.

If the intention of the DDI is that a closed-out worker insight means the close-out of the action items arising from that worker insight, it is no less problematic:

• Do all worker insights have to be accepted and acted on?
• If not all, who gets to decide which ones?
• Are rejected worker insights incorporated into or accounted for in the data/score?
• What if a worker insight takes 6 months to resolve? How does this affect the data/score?

Overarching this conversation of course is the notion of “closed-out”.

Anyone who has had anything to do with operational safety issues around audits, incident investigations, hazard inspections or any other process that generates action items will appreciate just how problematic and prone to manipulation the close-out of action items are.

The worker feedback process is also layered with a qualitative assessment of individuals perceptions of the process – what did people think about the worker insight process?

I do not think it is necessary to make too many observations about this part of the process, apart from the fact that any sort of qualitative analysis/survey process (think safety cultural safety climate) is subject to manipulation.

Given some of the vagueness around the mechanics of the worker insight process and the inherent capacity for manipulation within the process, it is difficult to see how a green DDI pressure gauge offers any more insight into the efficacy of safety management in an organisation, or its legal compliance, than a traditional green traffic light showing that all workers have completed the correct number of Take 5s, or that the correct percentage of corrective actions have been closed out in the preceding month.

Trade-offs and socio-economic bias

Even if we only take the worker insight element of the DDI, it represents a significant investment in time and resources by an organisation.

It seems self-evident that the process of worker insights described in the DDI could only be implemented by large organisations with resources to develop the processes necessary to implement the mechanics of worker insights.

Small and medium-sized enterprises simply do not have the resources or the frameworks necessary to deliver something as multifaceted as worker insights – they can talk to their workers and typically do, but the administration and bureaucracy of the worker insight program is a different step again.

Even for large organisations there will have to be some level of trade-off – something will have to be stopped to be replaced with, or provide resources for, a worker insight program.

The other problem is that most organisations with enough size and resources to implement a worker insight program will need to incorporate their contractors, who in many cases do most of the high risk/high hazard work, into the process.

Obviously, there are a range of practical considerations that will need to be considered, but which have not been described in the DDI. For example:

• Who is going to resource the worker insight process – the contractor or principal?
• What contractual allowances will be included to allow the contractor to participate in the process?
• Will a contractor’s engagement in a principal’s worker insight program increased legal risk for the principal under normal contractor safety management principles? For example, if a contractor requires a principal to change safety processes based on feedback from the worker insight program?

Over and above the technical implantation questions, in my experience many of the problems around the corruption of health and safety information gets magnified through the lens of a contractor. If contractors are prepared to “fudge” injury data, there seem to be far less impediments to manipulating all the various elements of a worker insight process.

The heart of the problem

While I think the technical concerns, trade-offs and economic problems should be enough for anybody to think seriously about adopting work insights as a safety measure, there is a genuine issue to be considered regarding the extent to which worker insights as a metric cannot meaningfully contribute to our understanding of safety efficacy or legal compliance.

While I think there are any number of very good reasons why organisations should engage with their workers around health and safety and seek their inputs, in my view, the collection of those inputs does not amount to an organisation doing everything reasonably practicable, nor to officers exercising due diligence.

On page 15 of the DDI, the following observation is made:

Assessments of the effectiveness of close-out of worker insights must also include consideration of whether those worker insights are leading to better health and safety performance in the organisation and having a positive impact on the organisation and its health and safety capacity. (my emphasis)

The DDI does not tell us how to measure whether worker insights are leading to better health and safety performance in the organisational or having a positive impact on health and safety capacity – indeed it seems arguable from the DDI that worker insights were supposed to do that.

Later, on page 15, the following observations made:

Organisations are encouraged to provide reporting commentary with respect to the processes for how they are assessing the effectiveness of closeout on worker insights. That is, how is the organisation and its leadership assuring themselves that the solutions identified by workers for better health and safety performance through worker insights are being implemented and are effective and fit for purpose. (my emphasis)

The DDI does not tell us how to measure whether implemented worker insights offer better health and safety performance or are effective and fit for purpose.

These observations go to concerns at the heart of worker insights, and indeed the entirety of the DDI – or any other safety performance metric.

At its core, the mechanics of worker insights is a measure of activity and closeout – like multiple existing health and safety measures – that reveal very little if anything about the efficacy of safety management and legal compliance.

If organisations could identify, measure, and articulate how safety initiatives are:

• leading to better health and safety performance;
• having a positive impact on the organisation and its health and safety capacity;
• are being implemented; and
• are fit for purpose,

there would be no need for a DDI.

As it is described in the DDI, the success of worker insights as a metric seems wholly dependent on an organisation’s ability to achieve the outcomes described above, without providing any advice on how to achieve those outcomes.

I am a company officer in my own right, and I am engaged on various risk and audit committees. In both of those capacities, I want to know whether my organisation has the right processes for managing the health and safety risks in the business and whether those processes are implemented and effective. Engaging with workers on health and safety might contribute to those outcomes, but a metric based on worker insights is not a measure of any of those things.

Final thoughts

There is an observation from a 2015 Queensland coroner’s inquest (Inquest into the death of Cameron Brandt Cole, 11 September 2015) which says:

The identification, elimination or minimisation of risks through risk management processes may lead to the production of a suite of documents that will pass audit requirements. However, the evidence at this inquest suggests that the workers in the field may find such documents hard to comprehend and of limited relevance to their daily activity (page 24).

Like so many health and safety metrics, the DDI appears to confuse the enablers or mechanics of effective safety with the outcomes.

In my experience, the purpose of engaging with workers around health and safety is to gain an insight into whether they understand our safety requirements, whether they are fit for purpose, and what we can do as an organisation to improve those processes – considering the workers perspectives.

Worker insights are an enabler of a safety outcome – they are not the outcome.

The reporting requirement is not how many times we have spoken to our workers, or how many new ideas they have come up with, or even how much they enjoyed the process. The reporting requirement to the organisation – particularly to satisfy legal requirements – is reporting about whether we have proper systems in place to manage health and safety risks in the business, and what, or what level of assurance we have about the extent to which those processes are implemented and effective.

Ultimately, safety performance is a narrative not a number and if history is any indication, any attempts to represent safety performance as a metric will fall into the same trap of measuring activity over assurance or enablers as outcomes.