This paper follows the rational political budget cycle approach, ,extending it in two directions. First, it considers an open economy, to formalize the implications of political budget cycles for nominal exchange rates. The variable used as a signal of competency is the rate of devaluation which, in the context of the one-sector model we use, coincides with the rate of inflation, and acts as a tax on consumption. The relevant trade-off is between devaluation today and tomorrow, as in the Sagent-Wallace unpleasant monetarist arithmetic. Hence, the pattern of devaluations around elections are part of a political budget cycle, a feature that has been overlooked in conventional stories of political budget cycles that concentrate on a closed economy. The second extension is more fundamental, and applies to political budget cycles in open and closed economies. The typical assumption in this class of models is that governments share the utility function of voters, but derive additional utility from being in office (which may lead to opportunistic behavior). The only informational asymmetry regards the degree of competence of the government. We introduce a second dimension over which there is incomplete information: the degree to which the government is self-motivated. Voters do not know whether the incumbent is opportunistic or not. This simple assumption turns out to have important implications.