Safe Man-Hour Milestones: How Companies in Asia Can Prove a Workplace Safety Record Is Real

Drive past a large construction site, factory or refinery in Malaysia, Indonesia or Singapore and there is a good chance you will see a board at the gate showing how many hours have been worked without a lost-time injury. The figure often runs into the millions, and it is displayed with pride. It may be fully deserved. But the same number can also describe a workplace where reporting an injury costs the team its bonus, where contractors sit outside the count, or where a worker with a fractured hand is moved to a desk so that no workday is technically lost.

That ambiguity is a real problem for sustainability managers, safety leaders and boards. Workplace health and safety is one of the few social topics where companies publish hard numbers that can be compared year on year. That makes the numbers valuable. It also makes it unavoidable to ask whether they can be trusted.

Why safety is the social metric that gets compared

The scale of the issue is large. The International Labour Organization estimates that nearly three million people die each year from work-related accidents and diseases. Asia and the Pacific account for around 63 per cent of that total, largely because of the size of the region’s workforce.

Disclosure expectations have followed. Bursa Malaysia’s enhanced sustainability reporting framework lists health and safety among the common sustainability matters for Main Market issuers. Its indicators cover work-related fatalities, the lost time incident rate (LTIR) and the number of employees trained on health and safety standards. Internationally, GRI 403: Occupational Health and Safety 2018 asks organisations to report fatalities, high-consequence injuries and recordable injuries as both numbers and rates. That applies to employees and to other workers whose work or workplace the organisation controls.

The legal backdrop has also changed. In Malaysia, the Occupational Safety and Health (Amendment) Act 2022 came into force on 1 June 2024. It made four main changes:

• it extended the Occupational Safety and Health Act 1994 to almost all workplaces;
• employers with five or more employees must now appoint a safety and health coordinator;
• the maximum fine for breaching core employer duties rose from RM50,000 to RM500,000;
• directors and certain officers can now be held liable for offences committed by the company.

A safety figure in a sustainability statement now sits alongside much sharper legal exposure.

What an injury rate does and does not tell you

GRI’s guidance makes a point that is easy to overlook. Injury data measure the harm that workers suffered, not how safe a workplace is. A low number is an outcome. On its own, it does not show whether hazards are under control or whether the organisation was simply fortunate during the reporting period.

Three technical issues make raw safety figures harder to compare than they appear:

• The multiplier. GRI allows rates to be expressed per 200,000 or per 1,000,000 hours worked, provided the company states which it uses. Identical performance can therefore appear as 0.5 in one report and 2.5 in another. Comparing peers without checking the base produces meaningless conclusions.
• The definition. Whether an injury counts as “lost time” depends on how an organisation handles modified or restricted duties. GRI defines a high-consequence injury by recovery time rather than lost time, partly because lost time reflects lost productivity as much as severity.
• The boundary. Counting or excluding contractors, agency staff, subcontractors’ workers or gig workers can move a rate substantially. Singapore’s national statistics show this clearly. In its 2025 report, the Ministry of Manpower recorded a major injury rate of 15.7 per 100,000 workers when platform workers were excluded. When they were included, the rate was 17.7. It was the first year their data had been collected, after the Platform Workers Act took effect.

Calculation errors happen even in listed companies with established reporting. Carlsberg Malaysia, for example, disclosed that its 2023 LTIR data had been restated because of a calculation discrepancy. A restatement is a sign that controls are working, not a scandal. But it shows why the method behind a safety number deserves as much attention as the number itself.

Five ways a safety milestone can mislead

Most misleading safety claims are not deliberate fabrications. They come from incentives and definitions that quietly shape what gets recorded. The table below sets out the most common patterns and the evidence that tends to expose them.

Pattern	What happens	Evidence that tests it
Reporting suppression	Rewards tied to injury-free periods discourage people from reporting injuries.	Near-miss and first-aid reporting trends; anonymous worker surveys; clinic and insurance records compared with the injury register.
Case reclassification	Injured workers are moved to restricted duties so that no day is recorded as lost.	Restricted-work and medical-treatment cases disclosed alongside lost-time injuries; return-to-work records.
Boundary exclusion	Contractor and subcontractor hours or injuries sit outside the headline figure.	Contractor hours and injuries reported separately; site access and permit-to-work records.
Inflated hours	Hours worked are estimated generously or include staff who are not on site.	Hours reconciled to time-and-attendance or payroll systems; a documented estimation method.
Selective counting	The counter starts after the last incident, or the claim covers only the best-performing site.	A clear start date, a scope statement and the full incident history for the period.

The first pattern has been studied most closely. In a 2012 review, the US Government Accountability Office noted that experts and industry officials believed rate-based incentive programmes may discourage workers from reporting injuries. These are programmes that reward low reported injury rates. The US Occupational Safety and Health Administration later said such programmes are acceptable only where employers take real precautions so that workers still feel free to report. Examples include rewarding hazard identification, training people on their reporting rights and checking whether people actually report. Regulation differs across Asia, but the behavioural lesson applies in any workplace.

The Safety Evidence Ladder

A practical way to judge any safety claim, whether your own or one in a peer’s report, is to ask how far up the following ladder the evidence reaches.

1. Stated. “We had zero lost-time injuries.” A claim with no visible method, scope or period.
2. Recorded. Incidents are logged in a register with dates, locations and classifications.
3. Defined and reconciled. Definitions of lost-time, recordable and high-consequence injuries are written down. Hours worked are reconciled to payroll or attendance systems. The rate base and the worker boundary are disclosed.
4. Corroborated. Leading indicators and independent records point the same way as the headline number. Near-miss reporting is healthy, high-potential incidents are investigated and corrective actions close on time.
5. Independently checked. Internal audit, an assurance provider or another qualified third party has tested the data and the controls behind it.

Most organisations sit at level two or three. The step from three to four matters most. At that point a low injury rate stops being an isolated figure and becomes consistent with how the operation actually behaves. Our guide to ESG assurance in Asia explains what level five can involve in practice.

Leading indicators that make a lagging number credible

Lagging indicators count harm after it has happened. Leading indicators track the activities and conditions that prevent it. A study of Malaysian offshore oil platforms found both kinds of link:

• lower recordable incident rates went with stronger management and worker engagement on safety and better management of change;
• lower lost-time injury rates went with stronger hazard identification and risk assessment.

For managers, the practical lesson is that the two kinds of indicator should move together.

Leading indicators worth tracking include:

• near-miss and hazard reports per 100 workers, treating a healthy volume as a positive sign rather than a problem;
• high-potential incidents identified and investigated, which GRI 403 recommends reporting even where no one was hurt;
• the percentage of corrective actions closed by their due date;
• checks that critical controls are in place for high-risk work such as work at height, lifting, confined spaces and energy isolation;
• completion of risk assessments, now an express statutory duty for employers in Malaysia;
• safety training coverage for employees and contractors, which also supports Bursa’s training indicator;
• worker participation, such as the share of safety committee recommendations that are acted on.

Take a site that reports a million injury-free hours alongside almost no near-miss reports and a backlog of overdue corrective actions. It deserves more scrutiny, not less.

A checklist before announcing a milestone

Before a safety milestone goes on a banner, into an annual report or into a press release, the responsible manager should be able to answer yes to each of the following:

• Is the period clearly defined, with a start date that was not chosen after the fact?
• Are contractors and other non-employee workers under the company’s control included, or is their exclusion stated explicitly?
• Are hours worked drawn from attendance or payroll systems rather than estimates?
• Are restricted-work and medical-treatment cases disclosed alongside lost-time injuries?
• Have any fatalities or serious injuries at the same operations during the period been disclosed, including those involving contractors?
• Does the incentive structure reward reporting and hazard identification rather than an absence of reports?
• Would the underlying records stand up to review by an auditor, a regulator or an injured worker’s lawyer?

Who should own the number

Because incentives can shape safety data, ownership matters. Operational managers whose bonuses depend on low injury rates should not be the only people who classify incidents. Many organisations give classification to a safety or occupational health function with its own reporting line. They also give the board, or a board committee, regular sight of both lagging and leading indicators.

Where executive pay includes safety targets, a design based only on LTIR can bring back the suppression risk described above. Our analysis of ESG-linked executive pay in Asia looks at how boards can build more balanced KPIs. It also helps to read safety figures in both absolute and rate terms, as discussed in our piece on absolute versus intensity sustainability metrics. A falling injury rate can hide a rising number of serious injuries when hours worked grow quickly.

When a safety achievement deserves recognition

None of this means safety milestones should not be celebrated. Consider a long injury-free run on a complex project, achieved with full contractor coverage, healthy reporting and independently checked data. It reflects real operational discipline and deserves attention inside and outside the organisation. The question is what kind of recognition fits.

Regulatory compliance, ISO 45001 certification and approvals from safety authorities answer a specific question: does a management system meet defined requirements? They are not interchangeable with public recognition of an outstanding result, and no form of public recognition replaces them.

Record recognition in Asia has often focused on scale and participation. Safety is a more demanding category. The evidence has to show that harm did not happen, and that no one was discouraged from reporting it when it did.

Some organisations with an exceptional and well-evidenced achievement explore independent recognition as a way to document it. Asia Record recognises measurable business achievements across Asia, and each proposal is assessed against a clearly defined claim and supporting evidence. Asia Record’s verification process reviews submissions for measurability, verifiability and ethical execution, among other criteria. For a safety milestone, the ethical test is central. An achievement built on suppressed reporting should not qualify for any form of recognition.

For companies considering how to get an Asia Record for a safety or wider ESG achievement, the Safety Evidence Ladder is a useful readiness test. Evidence at levels four and five is the kind of material that a record assessor, an investor and a regulator would all expect to see. That means leading indicators that support the headline number, plus independent checking of the data. Organisations that decide to apply for Asia Record recognition can review the published Asia Record application steps. They should treat any resulting corporate record as a complement to legal compliance and formal certification, never a substitute for either.

The number is only as good as the reporting culture

A safe man-hour milestone is one of the most visible social claims a company can make, on the site gate and in its sustainability statement. It is persuasive only when the organisation can show that people feel free to report, that everyone doing the work is counted, and that the hours behind the figure are real. Companies that build that evidence can report their safety performance with confidence.