- We Want To Study How Having Health Insurance Coverage Influence Medical Expense Using The Data From The Medical Expendit 1 (21.85 KiB) Viewed 90 times
1. However, there is a problem. The medical expenses of an individual apparently depend on whether or not they have health insurance. At the same time, it is more likely that people would choose to obtain health insurance if their medical expenses are high; hence whether an individual is covered by health insurance is also affected by the medical expenses. Because of the causality can go in both directions, the linear model above cannot be estimated by an OLS directly Which of the following best describes the problem as mentioned above? Heteroskedasticity Omitted Variable Bias Endogeneity Selection Bias 2. To get around this issue, one way is to use an instrumental variable. Here, we will use social security income replace ratio as the instrumental variable. What conditions below need to be satisfied by the social security income replacement ratio (ssiratio) for it to be a good instrumental variable? (Select all that apply.) The social security income replacement ratio is correlated with having health insurance coverage. The security income replacement ratio is correlated with control variables X The social security income replacement ratio is correlated with the error The social security income replacement ratio affects medical expense only (indirectly) through having health insurance coverage.