Public policies: will policy be ineffective if welfare is a random-walk

This study contradicts the view that policy cannot be effective if the welfare variable is a random-walk. Also, the study disagrees with the methodology employed by some proponents of this view, and goes on to show that the conclusion of policy ineffectiveness was only due to a wrong inference procedure. Furthermore, it points out that the notion of policy effectiveness suggests the examination of the relationship between policy and welfare; hence, what to look for is not a trend in the individual time series, but a common trend which relates the policy instrument and the welfare-variable in the policy environment under consideration.