Workplace accidents, a statistic that depends on reporting

French workplace accident statistics come from the data of the accidents at work and occupational diseases branch, that is from claims reported and then recognised.

That administrative origin determines their perimeter: it covers employees of the general scheme, with separate schemes for the public service and agriculture, which forbids a direct reading in terms of the total population. Self-employed workers, for whom cover is optional, are largely absent from it.

The system itself is old. The act of 1898 introduced flat-rate compensation without any finding of fault, in exchange for which the employee gave up full compensation. That compromise still structures the way claims are reported and recognised.

A commission established by law periodically assesses the cost to health insurance of claims not reported as workplace accidents or occupational diseases, and therefore borne by the health branch.

Its successive reports conclude that under-reporting is significant and give rise to an annual financial transfer between the two branches. The very existence of that transfer is institutional recognition of the phenomenon.

The causes put forward are many: ignorance of rights, fear for one's job, pressure linked to safety bonuses, or simple administrative convenience for short absences. Their respective weight is not measured precisely.

Recognition rests on tables listing pathologies, exposures and time limits. A disease not listed in a table may be recognised through a complementary procedure, longer and less frequent.

That system produces figures depending as much on the evolution of the tables as on that of actual exposures, which makes long series hard to interpret without knowing the regulatory changes. Pathologies with a long latency period, appearing decades after exposure, are the most affected by that lag.

Definitions of accident, severity thresholds and reporting rules differ between countries. European agencies publish harmonised data but flag persistent coverage gaps.

A country where reporting is encouraged and well covered may show a higher rate than one where it is less so, without the real situation being worse there. Fatal accidents, which are harder to leave unreported, are for that reason considered more comparable than accidents as a whole.

The long-term fall in the frequency of serious accidents in industry is robust and found in every country with long series. It reflects both prevention rules and the decline of industrial employment itself, two effects that aggregate series do not separate.

Recent developments are more mixed by sector, with growing attention to musculoskeletal disorders and psychosocial risks, whose recognition is more difficult and more recent.