Empirical Methods Final Exam: Correlation, Causation, and Treatment Effects

Correlation refers to a relationship between two variables, while causation implies that one variable directly influences the other. In empirical methods, establishing causation requires careful consideration of potential confounding factors and the use of experimental or quasi-experimental designs to infer causal relationships.

The key assumption in DiD is that the treated and control groups would have had similar time trends in the absence of treatment, known as 'common trends.' This assumption is essential for identifying the causal effect of the treatment.

The IV approach involves using an instrument (Z) that affects the endogenous regressor (X) to identify the causal effect of X on the dependent variable (Y). The assumptions required for IV validity include the instrument having a causal effect on X, independence assumption (Z being as good as randomly assigned), and exclusion restriction (Z affecting Y only through its effect on X).

The three needs that drive much of the financial activities of poor households are basic cash-flow management, coping with risk (including income and expenditure risks), and raising lump sums for big-ticket expenses.

Direct costs of risk refer to the variation in consumption caused by risk, while indirect costs of risk involve the costly actions individuals may take to reduce risk, such as raising lump sums for big-ticket expenses and making tough decisions.

Treated and control group would have had similar time trends in the absence of treatment

To address endogeneity in the relationship between the independent and dependent variables

Ensures that potential outcomes are orthogonal to treatment

'Z' has a causal effect on 'X'

Seek to eliminate selection bias in treatment assignment