RAPID: Health Decision-Making in the Aftermath of a Disaster: Evidence from Cyclone Idai in Mozambique
How do disasters affect individual decision-making about important health behaviors? Can public health interventions help reduce disaster-induced deterioration of individual health decision-making? We explore these questions in the context of Cyclone Idai in Mozambique, which affected an area covered by a US-government program to combat HIV/AIDS.
Exposure to a natural disaster can lead individuals to make worse decisions related to their health, potentially with consequences lasting beyond the period of disaster recovery. The countries most affected by the HIV/AIDS crisis (mostly in Sub-Saharan Africa) also have high exposure to weather, conflict, economic, and other shocks. Negative shocks are known to lead to increases in HIV infection (Oster 2012, Burke et al 2015), but the mechanism driving this relationship is not well understood. Disasters can delay HIV testing and interrupt antiretroviral therapy (ART), with lasting negative consequences: worse long-run outcomes for HIV infected individuals, increased risk of development of resistance to ART, and higher risk of HIV transmission to others. Higher rates of HIV transmission could be exacerbated by a rise in transactional sex in the wake of disasters (Dinkelman et al 2008). Delays in HIV testing, interruptions in ART treatment, and engagement in transactional sex could all be related to changes in individuals? economic preference parameters (e.g., related to time, risk, and motivated reasoning). For example, disasters could increase present-bias, leading individuals to place less priority on actions whose consequences occur farther in the future (such as HIV testing and treatment); increases in risk aversion, on the other hand, could have the opposite effect.
This study has three aims:
1. We will study how exposure to Cyclone Idai in Mozambique affects individual decisions related to HIV/AIDS: whether to have an HIV test; for HIV infected individuals, to adhere to antiretroviral therapy (ART); and risky sexual behaviors including transactional sex.
2. We will explore how changes in individual preferences and behaviors may mediate these effects. We will take measurements of risk preference, time preference, and motivated belief biases in our study population.
3. We will estimate whether an ongoing US-government HIV/AIDS program helps mediate the effects of the cyclone on HIV testing and treatment and on individual preference parameters.
- Dean C. Yang
2019-06-01 - 2020-05-31