Earnings and work accident risk: a panel data analysis on mining

Mining represents one of the most hazardous work environments. Health and safety standards tend to vary across countries, depending on the state of infrastructures, technological development, and exploration and development priorities within the sector. Following an overview of hypotheses on work accident exposure and related earnings differentials, the analysis examines the rationale for testing some of these hypotheses through alternative panel data models, by eventually accounting for unobserved heterogeneity, weak exogeneity and persistence. Regressions are estimated on an unbalanced panel of 12 industrial and developing economies. Factors related to the local economy and institutional background, are found to explain varying levels of the two dependent variables across time and space. Mine fatal injury exposure appears to decline with increasing levels of development and higher private investment rates. Earnings differentials with other sectors tend to widen with a large mining sector and high rates of unemployment. The hypothesis that differentials in gross earnings in mining relative to other economic sectors are also determined by increased fatal accident risk is not substantiated by econometric estimates.