I develop the implications for real exchange rate cycles of different policy preferences, focusing in particular on broadly stylized features of major Latin American and East Asian economies. Recent political science literature has emphasized the role of factors such as the influence of the manufacturing sector and the nature of labor markets. I formalize some of these insights in a developing country framework with policy makers who inter-temporally optimize and voters/audiences that are incompletely informed. Given the choice between assigning greater weight to immediate worker purchasing power versus generating manufacturing employment and income over time, I show that countries where policy makers choose the former are more likely to experience cycles with overvaluation, current account deficits, and abrupt (postponed) devaluations.