This special issue of the AJIM presents new evidence of undercounting of occupational injuries in the Bureau of Labor Statistics (BLS) Survey of Occupational Injuries and Illnesses (SOII). These studies review the history and current status of the SOII and illustrate the very real challenges to obtaining accurate data on occupational injuries. This commentary explores questions that inevitably follow from these findings: What underlies the underreporting within the SOII system? Why does it matter? What should be done about it? After prior research had shown that the much relied-upon BLS SOII data provide an incomplete and therefore potentially misleading picture of the incidence of occupational injuries and illnesses [Rosenman et al., 2006; Boden and Ozonoff, 2008], the Bureau of Labor Statistics funded additional studies. The initial results of this research are included in this special issue; additional multi-year studies are underway [Wiatrowski, 2014]. The papers presented here focus on three states (California, Massachusetts and Washington) and use three methodologies (case matching, capture-recapture and semi-structured interviews). The casematching studies draw from three data sources: the SOII, workers' compensation claims, and reports from physicians and hospitals. Three studies look specifically at amputations; of these, two use case-matching (in California and Massachusetts) and one uses capture-recapture methodology (in Massachusetts). [Appendix-Table I] The overall conclusion, across the three jurisdictions and across methodologies, is that the SOII significantly undercounts-and therefore underestimates-the number of injuries, even when looking only at objectively verifiable and often serious injuries such as amputations. The specific findings vary somewhat due to different methodological approaches and to variations among the jurisdictions in reporting requirements for both workers' compensation and health surveillance systems. All illustrate problems with the SOII, and they also demonstrate that no single known data source provides a complete picture of occupational injury. Before exploring these issues further, it is important to note that these problems do not suggest that the SOII is without value. To the extent that there is consistency in reporting over time, the SOII may accurately reflect trends such as the increasing share of cases that resulted in restricted work rather than time away from work [Wiatrowski, 2014]. But the underlying concern remains: researchers, regulators, legislators, workers, unions, insurers, and the public may be relying on data that significantly underestimate the magnitude of an important problem.
Emily A. Spieler J. D., Edwin W .Hadley Professor of Law, Northeastern University School of Law, Boston, Massachusetts 02115