| This dissertation is a study of policy coordination with an emphasis on the labor market and supply shocks. The simplest informational approach is used, which is the information exchange. As a first and crucial step to study coordination using this simplest approach, it is necessary to justify why information is scattered in the first place. The justification we have in this dissertation comes from recent business-cycle theory in an imperfect information world.; The problem is studied by means of a rational expectation model with signal extraction. One contribution of this dissertation besides the coordination issue is that we have an explicit solution for the model which is hard to get in a signal extraction model because it usually involves solving a polynomial equation.; By applying a similar framework, I also studied the coordination problem in a three-country framework. If a country's output depends largely on exporting intermediate materials to certain countries and its consumption goods come also from the same countries, like OPEC does, then coordination might be necessary.; If global coordination is impossible, it is proved that it is in the country's best interest to choose the country with the most unpredictable state of nature as its coordinating partner, because the other country's policy can be predicted with less error. In this dissertation, it is shown that even partial coordination (the coordination between the country and one of its trade partner) will lead to welfare improvements over the no-coordination solution.; Part Two of this dissertation discussed how to combine the exchange-rate policy and monetary policy together to achieve some targets. I use monetary policy in addition to exchange-rate policy as an example to show that an optimal combination of these two instruments allows the achievement of the two targets at the same time. Indeed, the set of weights chosen in a currency basket peg weights is just a simple weighted-average of another two sets of weights, each of which is optimal for a single target. |
