The under-reporting of injuries for Hispanic workers in construction

Using data from the Bureau of Labor Statistics and Current Population Survey, this work will examine under-reporting of nonfatal workplace injuries and illnesses in construction for Hispanics from 1976-2008 at the national level. Previous research implemented different methods and techniques to examine the reasons behind under-reporting of workplace injuries and illnesses, concluding that the current surveillance system suffers from a problem of chronic under-reporting. By focusing on Hispanics working in construction, and by using different methods to examine to what degree the presence of Hispanics at the worksite contributes to the under-reporting of nonfatal occupational injuries and illness, our findings suggest that the presence of Hispanic construction workers significantly reduces the number of reported injuries. A lagged fixed effect model will be used to investigate the under-reporting issue in construction with regard to Hispanics. Usage of this model constituted the first departure from previous research that studied under-reporting in which capture-recapture and face-to-face interviews were used. The second departure was implementing this method in the construction industry. This is the first time a lagged fixed effect model has been used to study this phenomenon (under-reporting in construction), while most literature about under-reporting has been restricted to manufacturing with some covering the overall economy. Lastly, this work is different in its duration and scope: It covers the years from 1976 up to 2008 and includes all 50 states and the District of Columbia. From a base model (without inclusion of Hispanics) we can see that the real workers’ compensation rate negatively impacted all types of injury reporting except light-duty, while union rates, unemployment, wages, and firm size correlate negatively with all types of injuries. When Hispanics and their interactions with union rate and workers’ compensation rates were included and summed as independent variables, we see that the presence of Hispanics leads to a decrease in the incident rate for all types of injuries except for light-duty. In other words, the presence of Hispanics leads to lower reporting and, therefore, the appearance of lower rates of all types of injury rates except for light duty which indicates that Hispanics are under-reported.