Oxford Bulletin of Economics and Statistics

I study the impact of stress resulting from the unexpected death of a child on parents' mortality risk. Using a flexible approach and allowing for time-varying treatment effects, I find no impact on the short-run mortality risk. However, I estimate a substantial increase in the long-run, particularly among men. I provide evidence that this group likely resorts to risky health behaviours such as heavy smoking and drinking as coping mechanisms. Assessing the possible protective effects of mental health support, I find evidence that it lowers the mortality risk for women. Mental health support has a smaller impact for men. I show that this is likely due to underutilization by those who would benefit the most. Finally, I present robustness of my results to specific departure from my identifying assumptions.