This paper attempts to make the case that a 2-sector model using the familiar traded non-traded distinction offers a reasonably successful empirical account of why Mexico needed to devalue its exchange rate in 1994. This model provides a way to define and measure disequilibrium in the exchange...