Labour Inspections and Workplace Safety: Evidence on Injury Prevention, Compliance and Effective Enforcement

Labour Inspections and Workplace Safety: Evidence on Injury Prevention, Compliance and Effective Enforcement

Labour inspections are an important part of occupational safety and health regulation, but their effect depends on the quality and consequences of the visit. This review, covering evidence from almost six decades, found that inspected workplaces had a 25 percent lower risk of work related injury than those not inspected. Inspections can reduce harm by identifying hazards, prompting corrective action and reinforcing the possibility of enforcement where employers fail to comply. Their effectiveness varies by sector, company size, inspection scope, inspector competence and follow-up support, with stronger results where practical advice is combined with credible enforcement. Evidence on compensation claims, sickness absence and occupational disease was less clear, partly because some conditions take years to develop. The review also warns that falling inspector numbers and rising workloads may reduce inspection coverage and quality.

Summary

The study concludes that labour inspections are associated with fewer work related injuries and better compliance with occupational safety and health requirements.

The meta-analysis found that inspected workplaces had an estimated 25 per cent lower risk of injury than non-inspected workplaces.

Inspections also appeared to reduce non-compliance, although the pooled compliance result was just outside the conventional threshold for statistical significance.

Evidence relating to compensation claims, occupational diseases, sickness absence and psychological or musculoskeletal outcomes remained limited or inconsistent.

The effectiveness of inspections depends on inspection design, company characteristics, inspector competence, follow-up consultation, enforcement orders, information sharing and the credibility of possible penalties.

The study calls for better-resourced inspectorates, harmonized data systems, longer-term research and inspection methods suited to changing work arrangements.

Aim & Context

Work-related accidents and diseases continue to create a substantial human, social and economic burden. Governments respond through occupational safety legislation, prevention programmes and enforcement systems intended to ensure that employers provide safe and healthy working conditions.

Labour inspections form a key part of this system. Inspectors may examine machinery, work processes, risk assessments, protective equipment, exposure controls, management arrangements and compliance with legal duties. They may provide guidance, require corrective action, issue formal orders or recommend sanctions.

Previous reviews had already indicated that workplace inspections could reduce injuries and improve compliance. However, much of the evidence covered studies published before 2017. Occupational safety systems, work organization and enforcement capacity have continued to change, making an update necessary.

The authors also identified an institutional evidence gap. In Germany, Austria and Switzerland, accident insurance organizations may inspect workplaces alongside government authorities. Research had concentrated mainly on state inspections, leaving uncertainty about whether inspections undertaken by insurance institutions produce comparable effects.

The study therefore asked: what effect do occupational safety and health inspectorate interventions have on safety and health within companies?

Its objectives were to summarize evidence published since the previous review, examine reports from authorities and insurance organizations, assess outcomes including injuries and legal compliance, and combine suitable studies in updated meta-analyses.

Methodology

The researchers conducted a scoping review with an additional quantitative meta-analysis. The project was commissioned by the German Federal Institute for Occupational Safety and Health and its protocol was registered in advance through the Open Science Framework.

The methodological approach was based on an earlier systematic review that covered evidence published between 1966 and February 2017. The new search focused on studies published from January 2017 onwards, allowing the researchers to update the earlier findings while retaining continuity in eligibility criteria and outcome definitions.

Eligibility Criteria

The review included working populations in countries classified as developed economies. This restriction was intended to improve comparability because legal frameworks, regulatory institutions and inspection systems can differ substantially between high-income, emerging and developing economies.

Eligible interventions included inspections carried out by government authorities or accident insurance institutions.

The outcomes of interest were broad and included:

• Fatal and non-fatal occupational injuries
• Workplace accidents
• Occupational diseases and illnesses
• Health complaints
• Sickness absence and return to work
• Employee safety behaviour
• Improvements in working conditions
• Introduction of safety measures
• Use of personal protective equipment
• Compliance with occupational safety legislation

Epidemiological observational studies, intervention studies and systematic reviews were eligible. Qualitative studies, individual case reports, case series, narrative reviews, editorials, commentaries and publications with insufficient methodological information were excluded.

Study Selection

After duplicate records were removed, 5,498 publications from the electronic databases entered title and abstract screening. Ninety-seven database records were assessed in full text, resulting in the inclusion of 26 publications.

A further 92 publications identified through reference checking, citation searching and contacts with relevant organizations were assessed. Eleven of these were included.

The completed review contained 37 publications: 31 publications reporting epidemiological or intervention research and six publications relating to systematic reviews.

Most screening was conducted by one reviewer. A second reviewer independently assessed 20 per cent of titles, abstracts and full texts. Agreement was substantial, with a Cohen’s kappa value of 0.78. Disagreements were resolved through discussion.

Data extraction was also led by one reviewer, with a second reviewer checking 20 per cent of the extracted material for completeness and plausibility. Information collected included study design, country, occupational groups, inspection characteristics, comparison groups, outcomes, statistical methods, funding and conflicts of interest.

Meta-Analysis

Although meta-analysis is not always performed as part of a scoping review, the researchers used it to provide an updated quantitative estimate.

Suitable studies from the new review were combined with eligible studies from the earlier systematic review. Where several publications used the same database, time period and outcome, the most recent or comprehensive publication was selected to avoid double counting.

Separate analyses were conducted for occupational injuries and non-compliance. Compensation claims were not pooled because studies combined acute injuries with diseases and disorders that have very different latency periods.

Random effects models were used because variation between studies was expected. Two level or three level models were selected depending on whether studies reported several related effect estimates. The analyses were completed in R using the metafor package and restricted maximum likelihood estimation.

The review did not conduct a formal risk of bias assessment, GRADE evaluation or publication-bias analysis. These decisions were consistent with its scoping-review design but limit the certainty that can be assigned to the pooled results.

Key Findings

1. Inspections Were Associated With Fewer Occupational Injuries

Most individual studies examining injury outcomes found that inspections were associated with reductions in fatal or non-fatal workplace injuries.

The injury meta-analysis included ten publications retained from the earlier review and two publications identified in the update. These studies covered sectors including manufacturing, coal mining, transport and workplaces selected without a specific occupational restriction.

The pooled relative risk was 0.75, with a 95 per cent confidence interval of 0.57 to 0.97. This means that the inspected workplaces had an estimated injury risk 25 per cent lower than the risk in workplaces that were not inspected.

The confidence interval did not cross 1.00, indicating a statistically significant pooled association. The direction of the result was also consistent with the general conclusions of earlier systematic reviews.

Newer studies included in the update produced a smaller but more precise pooled effect, with a relative risk of 0.89. This still suggested a protective association, corresponding to an estimated 11 per cent lower risk.

These findings strengthen the case for inspection as a preventive measure. However, the result should not be interpreted as proof that every inspection automatically reduces injuries by the same amount. Inspection programmes differ in scope, intensity, follow-up and enforcement authority.

2. Study Heterogeneity Was Extremely High

The injury analysis had an I² value of 99.28 per cent. This indicates very high variation between the study results that cannot be explained by chance alone.

The studies covered different countries, industries, periods, legal systems, injury definitions and inspection models. Some inspections followed an accident, some were planned, some involved penalties and others emphasized support or consultation.

Because of this variation, the pooled figure provides an overall direction of effect rather than a universal prediction for individual workplaces. The result supports the protective role of inspections, but policymakers should consider the local legal, institutional and industrial context when applying it.

3. Inspections Appeared to Improve Legal Compliance

Five studies were included in the compliance meta-analysis: four from the previous review and one identified in the update.

The pooled relative risk for noncompliance was 0.68. This suggests that inspected workplaces had an estimated 32 per cent lower likelihood of non-compliance than non-inspected workplaces.

The 95 per cent confidence interval ranged from 0.45 to 1.01. Because the upper limit slightly exceeded 1.00, the result narrowly missed the standard threshold for statistical significance.

The estimate remains practically important, particularly because several individual studies found improvements in the introduction of required measures and compliance with legal standards after inspection. However, the compliance analysis had exceptionally high heterogeneity, with an I² value of 99.95 per cent. The precise size of the effect is therefore uncertain.

4. Evidence on Compensation Claims Was Inconsistent

Results concerning workers’ compensation claims varied by industry, follow-up period and outcome definition.

Some studies found reductions in claims only after repeated inspections or only within particular sectors. One transportation study reported lower claims in the second year when more than one inspection had occurred. Construction studies sometimes found an initial increase, while findings in manufacturing were neutral.

An increase in recorded claims following inspection does not necessarily indicate that safety worsened. Inspections may improve awareness, reporting and access to compensation, temporarily raising the number of documented cases.

The studies also combined acute injuries with musculoskeletal disorders and other conditions that may take years to develop. This made it difficult to determine whether an inspection had influenced the underlying occupational disease process.

Longer follow-up periods and clearer separation of injury and disease outcomes are needed.

5. Evidence on Occupational Disease Was Particularly Limited

None of the included primary studies directly examined the effect of inspection on recognized occupational diseases as a main outcome.

One randomized controlled trial examined sickness absence related to musculoskeletal and psychological diagnoses. It found a non-significant tendency toward fewer musculoskeletal-related sickness days but an increase in absence related to psychological diagnoses during the study period.

The follow up lasted approximately 18 months, which may have been too short to detect meaningful changes in conditions with long latency periods.

Inspections may reduce exposure to noise, hazardous substances, poor ergonomics or other chronic risks, but their health effects may not appear in compensation or disease records for several years. Future studies need extended observation periods and access to linked inspection, exposure and health data.

6. Company Characteristics Influence Inspection Outcomes

The effectiveness of inspection was affected by company size, industry, previous safety experience and the way workplaces were selected.

High risk industries may benefit from targeted inspection because their potential for serious harm is greater. Targeting can also make limited inspectorate resources more efficient.

Concentrating exclusively on workplaces above a numerical risk threshold can create unintended gaps. Companies just below the threshold may receive little attention despite having significant hazards. A balanced model may therefore include high-risk establishments, workplaces with rising incident rates, low-rate comparison establishments and companies that fail to provide required information.

Larger companies may record more violations simply because their operations are more complex and inspectors have more activities to examine. They may nevertheless have stronger internal occupational safety systems than small firms. The number of recorded violations should therefore be interpreted alongside workplace size and inspection scope.

7. Credible Enforcement Can Strengthen Compliance

Inspections involving formal orders were often more effective in prompting risk-control measures and reducing injuries than visits without enforceable requirements.

Higher penalties also appeared more influential than minor penalties. The expected consequences of continued non-compliance can change employer behaviour by making inaction more costly.

The findings do not support applying severe penalties in every situation. Formal sanctions may be most appropriate where serious risks exist, violations are repeated or an employer fails to respond to earlier advice or written requirements.

An effective enforcement system needs proportionality. It should distinguish between organizations willing to correct deficiencies and those that repeatedly disregard legal duties.

8. Consultation Can Increase the Value of Inspection

Several studies indicated that inspections followed by consultation with managers were more successful in achieving workplace improvements than inspection alone.

Consultation helps employers understand why a requirement exists, how a hazard should be controlled and what acceptable implementation looks like. This is especially relevant for small organizations that may lack specialist occupational safety expertise.

The evidence suggests that consultation and enforcement should not be treated as competing approaches. Guidance can improve understanding, while credible sanctions ensure that employers cannot disregard serious or repeated violations.

A combined approach may offer the strongest balance: inspectors explain the required changes, provide reasonable opportunities for correction and retain the authority to issue orders or penalties when action does not follow.

9. Inspector Knowledge and Communication Matter

Inspectors’ education, professional background and ability to explain necessary changes can influence the result of a workplace visit.

An inspector must identify hazards, interpret regulation, assess the adequacy of controls and communicate findings in a way that employers and workers can apply. A technically correct inspection may have limited impact when the required changes are unclear or poorly explained.

Differences in inspector training can also change which hazards are prioritized. Some inspectors may focus on immediately visible violations, while others may examine management systems, psychosocial conditions or less visible long-term health risks.

The study recommends more standardized strategies for monitoring and enforcing occupational safety law. Greater consistency would support inspection quality and make results more comparable across regions and countries.

10. Information Sharing Is Essential

Safety improvements at one company site do not necessarily spread to other sites belonging to the same organization.

A business may correct deficiencies in the inspected location while similar risks remain elsewhere. Differences in regional enforcement or poor communication between authorities can also allow companies to shift attention toward areas where scrutiny is stronger.

Sharing inspection and violation information within and between authorities may reduce this fragmentation. National coordination is especially important for organizations operating across several regions.

Authorities also require access to reliable, anonymized data on inspections, violations, corrective action and injury outcomes. Without linked longitudinal information, it is difficult to determine which inspection models produce durable improvement.

11. No Evidence Was Obtained on Inspections by Accident Insurance Organizations

The researchers specifically sought evidence on inspections conducted by accident insurance institutions. Relevant organizations were contacted, but reports were not supplied, largely because of data-protection concerns.

The final results therefore relate exclusively to inspections by government authorities.

This is an important limitation because accident insurers play a substantial inspection and prevention role in some European countries. Their activities may differ from government enforcement in purpose, legal authority, technical support and relationships with employers.

The absence of accessible data prevents an evidence-based comparison between government and insurance-based inspection systems.

12. Inspectorate Capacity Is Under Pressure

The paper reports declining numbers of inspectors and workplace visits in Germany and similar pressures internationally.

In Germany, the number of state labour inspectors fell from approximately 3,600 in 1991 to 1,620 in 2023. Company inspections by state inspectorates declined from more than 624,000 to around 134,000 during the same period.

Inspectors have also taken on responsibilities outside occupational safety, reducing the time available for workplace health and safety activity. Larger geographic areas, time pressure, insufficient resources and exposure to harassment can contribute to stress and burnout among inspection staff.

Reduced staffing has two consequences. Fewer workplaces can be visited, and individual inspections may receive less time for preparation, investigation, consultation and follow up.

An inspection system cannot produce its full preventive effect without sufficient numbers of properly trained and supported inspectors.

13. Traditional Inspection Methods Must Adapt to New Work Arrangements

Most of the primary studies used data collected before the COVID-19 pandemic.

Remote work, flexible schedules, subcontracting and decentralized workplaces can make hazards less visible and reduce the usefulness of inspection approaches designed mainly for fixed industrial sites.

Inspectors may need new methods for evaluating home-based work, digital management systems, dispersed workforces and psychosocial risks. Digital records and remote monitoring may support this process, but they cannot always replace direct observation and conversations with workers.

The study calls for adaptive inspection strategies and updated research reflecting post-pandemic forms of work.

14. The Evidence Is Supportive but Not Definitive

The review has several methodological limitations.

Only one randomized controlled trial was identified, while most primary evidence came from observational studies. Observational findings can be affected by selection. High risk workplaces are more likely to be inspected, and differences between inspected and non-inspected companies may not be fully captured in statistical adjustment.

The review did not conduct a formal risk of bias assessment, GRADE evaluation or publication-bias test. Both meta-analyses also showed heterogeneity close to 100 per cent.

Most studies originated in the United States. Their findings cannot be transferred automatically to countries with different legal systems, inspector powers, labour markets or regulatory traditions.

The results should therefore be understood as strong supporting evidence of a protective association rather than a precise universal estimate of causal effect.

Takeaways for Practice

• Occupational safety authorities should preserve labour inspection as a central prevention function and provide inspectorates with adequate staffing, training and technical resources.
• Inspection programmes should prioritize sectors and workplaces with serious or increasing risks while maintaining enough broader coverage to avoid leaving near-threshold companies unexamined.
• Workplace visits should combine clear explanation with credible enforcement. Consultation can help willing employers implement improvements, while orders and meaningful penalties remain necessary for serious or repeated non-compliance.
• Inspectors should receive standardized guidance covering investigation, risk assessment, communication, documentation and follow-up. Training should include emerging work arrangements and long-term health risks as well as traditional accident hazards.
• Authorities should share information across regions and inspect multiple sites where company-wide risks are suspected. Corrective action at one establishment should not be assumed to have reached the rest of the organization.
• Inspection outcomes should be monitored over several years. Immediate measures such as corrected violations should be combined with longer-term indicators including injuries, exposure reduction, sickness absence and occupational disease.
• Governments and insurance organizations should make anonymized inspection and outcome data available for independent research while protecting workers’ and employers’ legitimate privacy interests.

Read the full research study here:

Scoping review on the effect of labour inspections on occupational health and safety: a meta-analytic update Melanie Schubert1*, Ulrich Bolm-Audorff1 , Johan Hviid Andersen2 , Gabriela Petereit-Haack3 , David Reissig1 and Andreas Seidler https://link.springer.com/content/pdf/10.1186/s12995-026-00497-6.pdf