Modifying an accident process and its justice system – From single narratives and retribution to multiple stories and restoration

This study from Derek Heraghty , Drew Rae and Sidney Dekker explored the design, introduction and sunsetting of a restorative justice process in a construction joint venture arrangement of two companies.

Providing background:

·         “There have been a number of arguments put forward on the benefits of using restorative justice mechanisms when responding to workplace accidents”

·         Some findings “have even shown the positive economic effects created where a restorative-style just culture has been introduced”

·         “The most widely used accident analysis tools are based on sequential, reductionist models of systems and causality … Humans are often the target of such causal attribution, particularly due to outcome bias”

·         And “This can result in accident analyses halting at the point where they assume that the human involved had complete freedom of action”

·         Further, “Blame can significantly reduce organisational learning”

·         Just culture was introduced as a “means of reducing blame and increasing learning in industry. Its introduction was intended to strike a balance between no-blame and accountability”

·         Nevertheless, “Anxiety had emerged due to the introduction of no blame-systems … as organisations feared reckless actions would become immune from accountability”

·         Based on models created by Reason and Marx in just culture, while pioneering, “rely on a reductionist-style accident analysis, which have trouble acknowledging complexity”

·         “For a person’s actions to be passed through the culpability model, the accident must first be broken down into a chain of events. This oversimplifies the accident and excludes the system factors and the non-linear aspects of the event”

·         Second, “retrospective accountability, which the Reason culpability model unintentionally allows for, is susceptible to hindsight bias” and terms like “reckless violation” is “only ever placed on an action after something bad has resulted”

·         This retrospective accountability after the fact “can be seen as unfair as the judgment is based on the outcome rather than the action itself”

·         “Restorative justice is a process “where all stakeholders affected by an injustice have an opportunity to discuss how they have been affected by the injustice and to decide what should be done to repair the harm”

·         It’s said that a common misconception with restorative justice is that it’s the “opposite of retributive justice (Daly, 2002). On the contrary, many of the outcomes which emerge from a restorative approach to justice could be legitimately perceived as punitive by those involved”

·         “Retributive mechanisms are intended to answer pain with pain, whereas restorative mechanisms are used to work towards constructive outcomes”

·         Two types of accountability are described: retrospective accountability and prospective accountability

·         Whereas “retrospective accountability focusses on ensuring individuals pay their due for their wrong”, prospective “establishes the actor’s accountabilities for future action”

·         “ Prospective accountability holds people to account, not through blame, but by ensuring the actors involved provide their account”

The paper discusses the restorative justice approach which was introduced into the joint venture. I’ve skipped most of the initiative – so check out the paper:

1.       Accident process modification: An accident analysis process “infused with restorative justice mechanisms was introduced to the organisation to replace the old process which contained retributive justice mechanisms”

2.       Removal of suspension and punishment from the process: Suspension and other forms of punishment were removed from the process as they “were shown to cause psychological damage and negatively impact future safety”

3.       An adjustment of the language used to describe the process: The paper notes that “The language used as part of the accident process has been shown to influence both participation (Heraghty et al.,  ) and decision making” and hence, “Much of the language shown to have negative connotations was removed”

4.       Acceptance of multiple stories and moving away from reductionism: Whereas the original approach to investigations was a linear and reductionist one-dimensional story of root causes, the revised process included multiple narratives

5.       Inclusion of the workforce in the creation of actions

Note: I’ve skipped all of the detailed and specific results and gone straight to the discussion, so check out the full paper.

Results

Overall they found:

·         “Using a restorative justice-influenced accident process has benefits for the worker and the organisation”

·         “restorative justice mechanisms can improve honesty, engagement and learning”

·         “Accident learning approaches have universal requirements for success”, which include:

o   Positive treatment of the participants,

o   Transparency and consistency of the process,

o   Acknowledgement of participants’ experiences with previous accident

o   learning systems

·         “A restorative accident process is just as vulnerable to external forces and beliefs as any other process”

Unpacking some of the findings, it’s said that use of a restorative justice accident process has benefits for the worker and organisation. Removal of punishment and the empowerment of people involved in an accident were “viewed by both the workers and the management as crucial in improving the organisation’s learning culture”.

Some of the process changes in the initiative were said to have positive implications for participant wellbeing and their trust in the system. For instance, when this process was followed and individuals were informed that no punishment would be used “participants were significantly more willing to share their story with honesty”.

This contrasted with the previous accident process where individuals were “inclined to avoid the process entirely, if possible”. They cite other research highlighting that “punishment stifles learning, and its removal goes a long way in encouraging those involved in accidents to share their story”.

With workers more willing to share information and participate in the accident process, the organisation is said to benefit. This includes higher trust between those involved in accidents and the organisation, also believed to lead to better quality solutions resulting from the investigation; something “which the use of punishment can stifle”.

Role of language

They next discuss the vital role of language used in the accident process. That is, “Much of the negative perceptions of the previous accident process were initiated by the language used to describe it”. People involved in past investigations were left with the impression that they “had committed a crime and were on trial due to its similarity to the language used by law enforcement and the legal system”.  Hence, the language change involved in the restorative justice initiative was “deemed an important element in encouraging people to participate”

Role of communication

Moreover, the standard of communication following an event was also found to have a considerable effect on the treatment of people and actions. That is, “Where the swift and factual communication of an event was found to be lacking, those involved were more vulnerable to ill treatment and blame and more likely to leave the process with a negative perception of the system”.

Poor communication has also been linked to an organisation’s inability to learn. Despite best intentions, a “stakeholders’ hunger for immediate learnings from serious events can often result in inaccurate information being provided to satisfy them”. So, the rush to pull together initial or interim reports can “cause more harm than good as those reading are focussed on fixing the issues as quickly as possible and are likely to accept the information as final”.

They argue that a good communication system isn’t one that attempts to provide immediate answers to everything, but rather “one which ensures all involved are kept up to date on the process”.

They also discuss the role of how leaders and managers respond to incidents affects learning. When a person is “fearful of providing their manager with bad news [they] will likely hide all issues”.

Another barrier in the accident process is a person’s previous experience in the investigation process. People who have “faced or witnessed ill treatment in the past are more difficult to engage; regardless of the type of process”.

Development of restorative mechanisms

They next discuss mechanisms, saying that these should be part of a larger system:

·         Positive language and framing,

·         Management trained to treat their team in a compassionate manner following an accident

·         Transparency for all involved through all stages of the process.

A restorative accident process is just as vulnerable to external forces and beliefs as any other process

Here they argue that while there were many benefits to this approach, the approach is still susceptible to the same external threats as any other processes.

For instance, non-safety goals associated with the accident process are said to be a “significant risk”. Safety is often used as a part of brand and marketing/tendering, and therefore “Those who take an aggressive approach to accident management can be labelled as leaders, as the organisation can use their employee’s approach to show their stakeholders that their people take safety seriously”.

Problematically, “This can occur even when the aggressive approach taken is damaging to both safety and learning”. And when an organisation rewards this type of aggressive behaviour, “it can also lead to individuals using safety to exert power over others under the guise of taking safety seriously”.

Also notably – the paper discusses how the initiative was curtailed and sunsetted, but I’ve skipped most of this. However, the initiative was terminated in the 8th month of the 12-month trial, triggered by several factors. One was that the client was “not in the same space” regarding the new approach; also the client’s H&S team “showed a reluctance to move away from the ideals of Reason’s culpability model, particularly the ability to issue punishment where they, in hindsight, deemed an action as “reckless”.

Link in comments.

Authors: Heraghty, D., Dekker, S. W., & Rae, A. (2021). Modifying an accident process and its justice system–From single narratives and retribution to multiple stories and restoration. Safety science, 139, 105248.