Paul Bouscasse (Columbia University)

Abstract: Does an exchange rate depreciation depress trading partners' output? I address this question through the lens of a classic episode: the currency devaluations of the 1930s. From 1931 to 1936, many of the biggest economies in the world successively left the gold standard or devalued, leading to a depreciation of their currency by more than 30% against gold. In theory, the effect is ambiguous for countries that did not devalue: expenditure switching can lower their output, but the monetary stimulus to demand might raise it. I use cross-sectional evidence to discipline the strength of these two mechanisms in a multi-country model. This evidence comes in two forms: (i) causal inference of the effect of devaluation on country-level variables, (ii) new product-level data to estimate parameters that are essential to discipline the response of trade — the international elasticity of substitution among foreign varieties, and the pass-through of the exchange rate to international prices. Contrary to the popular narrative in modern policy debates, devaluation did not dramatically lower the output of trading partners in this context. The expenditure switching effect was mostly offset by the monetary stimulus to foreign demand.