I have a dataset with yearly levels of corruption in a number of countries. year, corruption 2001, 5 2002, 7 etc.
I want to test whether corruption is affected by a change in power (defined as the election of a new president who isn't part of the same political party as the previous one).
My take at this is to look at the slope in the corruption measure (∂corruption / ∂year) two years before and after the new president takes office. So any change change in slope after the new president (good or bad) is treated as a change. It's likely that some presidents will increase corruption, others reduce it and some will have no impact.
Any thoughts on a good statistical method to test this?