Some U.S. states enact laws aimed at reducing corruption by public sector officials, e:g: laws to protect whistleblowers" who reveal official malfeasance or misconduct. A researcher is interested in the effect of these anti-corruption laws on public sector corruption. He gathers data from all U.S. states in a particular year on (1) whether the state had enacted an anti-corruption law, and (2) the extent of public sector corruption in that state for that year. He then studies the relationship between these variables statistically.
(a) Is this research design cross-sectional, time-series, or some other kind? What is the unit of analysis?
(b) Suppose the researcher finds that states with anti-corruption laws have less public sector corruption than states without anti-corruption laws. Does this demonstrate that anticorruption laws are effective? Provide an alternative explanation that would create the observed relationship even if the laws have no effect at all.
(c) Suppose the researcher finds that states with anti-corruption laws have more public sector corruption than states without anti-corruption laws. Should the researcher write up a splashy paper arguing that anti-corruption laws have the opposite effect of the one intended? Provide an alternative explanation that would create the observed relationship even if the laws do in fact cause a reduction in public sector corruption.
(d) If you had a time machine and dictatorial powers over state law, how would you arrange an ideal research design to study whether anti-corruption laws cause a change in public sector corruption at the state level?
The question belongs to Statistics and it discusses about statistical significance of laws for reducing corruption in public sector organizations in the US.
Total Word Count 228Download Full Solution