labor · digital · family: the regulator demands evidence that cannot exist
injured on the job,invisible to the system
Gig and platform workers have no occupational injury reporting pathway because Workers' compensation is tied to employment status
Problem statement
An estimated 55 million Americans (35% of the workforce) engage in some form of gig or platform work — delivery drivers, ride-share operators, warehouse pickers, freelance tradespeople, home care aides. When these workers are injured on the job, there is no systematic reporting mechanism. OSHA's injury and illness recordkeeping requirements (29 CFR 1904) apply only to employers with employees — platform companies that classify workers as independent contractors have no legal obligation to record or report workplace injuries. Workers' compensation systems, the primary mechanism for tracking occupational injuries and financing treatment, similarly exclude independent contractors in most states. The result is a large and growing workforce whose occupational injuries are statistically invisible: they appear in emergency department records as "accidents" rather than occupational injuries, are not linked to specific work activities, and cannot be analyzed for prevention.
Why this matters
BLS data show that delivery drivers and warehouse workers have injury rates 2–3× the national average — but these rates are calculated only from employer-reported data and exclude the gig/platform segment entirely. Estimates suggest 50,000–200,000 serious gig worker injuries annually go unreported in the U.S. alone. Without data, there can be no evidence-based safety interventions: we cannot identify which tasks, routes, time pressures, or platform design decisions cause injuries. The growth of platform work means this data gap is widening annually, not narrowing. Globally, the ILO estimates 1 billion gig/informal workers with essentially no occupational injury surveillance.
What’s been tried and why it hasn’t worked
California's AB5 and similar laws attempted to reclassify gig workers as employees (which would trigger OSHA recordkeeping), but Proposition 22 reversed this for app-based workers, and the classification debate remains unresolved. Some platforms (Uber, Lyft) offer voluntary injury insurance, but these are claims systems, not epidemiological surveillance — they capture who filed a claim, not the universe of injuries. Academic surveys have attempted to estimate gig worker injury rates through self-report, but response rates are low (10–20%), selection bias is severe, and no survey captures the task-level detail needed for prevention. The BLS Survey of Occupational Injuries and Illnesses (SOII) explicitly excludes the self-employed. The fundamental barrier is structural: the entire U.S. occupational health data infrastructure was built on the employer-employee relationship, and platform work has disaggregated that relationship.
What would unlock progress
A worker-centered injury reporting system that is not mediated by the employment relationship. This could be: (1) a voluntary digital reporting tool (app-based, with appropriate incentives) that gig workers use to report injuries, with standardized taxonomy and geolocation linking injuries to specific work tasks; (2) emergency department intake protocols that identify and code gig work injuries separately from general "accident" codes; or (3) platform-level data sharing mandates where companies report anonymized injury-related trip/task data to OSHA or NIOSH. The key design challenge is incentive alignment — workers fear retaliation or loss of platform access if they report injuries, and platforms have no regulatory incentive to collect data that might demonstrate their work model is hazardous.
Entry points for student teams
A team could design and pilot a mobile injury reporting app for gig workers in a specific sector (e.g., food delivery), recruiting through worker centers and driver forums and testing usability, reporting rates and data quality against participants' own recall on a follow-up injury-recall instrument rather than against emergency department records — ED records are protected health information and linking them to platform activity requires a health-system data use agreement no student team will hold, so recall agreement, item completeness and drop-off across the reporting flow are the quality measures actually available. An epidemiology team could estimate the hidden burden from data that is already public: NIOSH's Work-RISQS query system (https://wwwn.cdc.gov/Wisards/workrisqs/) returns national estimates of ED-treated nonfatal occupational injuries from the NEISS-Work probability sample for 1998–2023 and, unlike the BLS SOII, explicitly covers self-employed workers alongside private-industry and government workers — enough to chart the delivery-and-transport injury curve and, more usefully, to document exactly where NEISS-Work's own occupational coding stops being able to see platform work, which is the measurement finding this brief is really about. Where discharge-level detail is needed, AHRQ's Nationwide Emergency Department Sample is a purchase, not a download: it carries a fee, a data use agreement and a required online training course through the HCUP Central Distributor (https://hcup-us.ahrq.gov/nedsoverview.jsp), so plan it as a gated resource with lead time, while AHRQ's free HCUPnet tool (https://datatools.ahrq.gov/hcupnet/) serves the aggregate case with no agreement at all. A policy team could design a regulatory framework for platform-level occupational injury reporting that is compatible with independent contractor status. Relevant disciplines: public health, epidemiology, human-computer interaction, labor policy, data science.
Genome — every gene is a door
Structural cousins — same reason stuck, other fields
Sources
ILO, "World Employment and Social Outlook: The Role of Digital Labour Platforms," 2021; OSHA, "Injury and Illness Recordkeeping — Coverage," 29 CFR 1904; Bajwa et al., "The Health of Workers in the Global Gig Economy," *Globalization and Health*, 2018; NAS, "The Design of the Current Employment Statistics Survey," 2023. Accessed 2026-02-25.
verification notes (working record)
The collection team’s own sourcing notes for this brief, kept verbatim:
Worsening mechanism: (1) the gig/platform workforce is growing at 10–15% annually, expanding the unmonitored population; (2) platform algorithmic management is intensifying work pace (documented by NIOSH), likely increasing injury rates; (3) the classification debate (employee vs. contractor) is moving toward contractor status in most jurisdictions, further excluding gig workers from occupational health systems. The data gap is getting wider as the workforce grows. Related briefs: labor-heat-stress-informal-agricultural-workers (similar pattern of occupational health systems designed for formal workplaces), education-skills-taxonomy-interoperability (workforce data system gap).
Reconciliation 2026-08-21: Entry-point repair under the C37 realism standards. The triage flag held: both data-bearing doors ran through emergency department records — validating the app "against ED visit records" and linking ED records to gig activity by temporal pattern — and ED records are protected health information requiring a health-system data use agreement, leaving only the policy door reachable. The app door now validates against a follow-up recall instrument, and the epidemiology door moves onto public data. Resources verified by direct fetch on 2026-08-21: NIOSH Work-RISQS, https://wwwn.cdc.gov/Wisards/workrisqs/ — free interactive query on NEISS-Work, 1998–2023, case definition confirmed on the page as "All workers (self-employed, private industry, and government)," with the note that NIOSH stopped collecting after 2023 and that case criteria changed in 2015; HCUP NEDS overview, https://hcup-us.ahrq.gov/nedsoverview.jsp — confirmed as a purchase ("Purchase the AHRQ HCUP NEDS"), roughly 30 million unweighted ED visits per year, and therefore labeled gated here rather than public; HCUPnet, https://datatools.ahrq.gov/hcupnet/ — live free query tool, named only for the aggregate case. Work-RISQS's inclusion of the self-employed is the substantive find: the national ED-based system that does cover this population exists, and the brief's real gap is that its occupational coding cannot identify platform work. Genome Tags untouched.