## Problem 206 — rank:243

**Title:** Causal effect of interventions

**Added binding assumptions (LaTeX):**

```latex
Specify the full cluster or network exposure intervention when units interfere, and formulate exchangeability and positivity for that exposure history rather than only a unit's own policy indicator. Define the unmeasured-confounding sensitivity model mathematically, including its range and time dependence. Identification by sequential adjustment is an established conditional result; estimating a credible policy contrast under violations is the residual research task.
```

**Direct edits:** 0. **Fresh external checks:** None; supplied sources retained as provenance.