Now let's think about a different context (ignore parts (a-c) above). Suppose that the government is concerned that local governments are expropriating transfers that are meant for local public goods provision. It turns out that an anti-corruption campaign was waged in Northern districts but not in Southern districts. The government has data on local public goods expenditure for each district before and after the anti-corruption campaign. Please (i) explain in words the difference-in-difference strategy you could use to estimate the effect of the anticorruption program; (ii) write down the actual DD estimator, defining your variables clearly; and (iii) explain the identifying assumption needed in order for this estimate to be valid