The Anti-Poor Distributional Effects Of Large Exchange Rate Devaluations

by Javier Cravino and Andrei Levchenko, Voxeu.org

Large exchange rate swings remain a prominent and recurring feature of the world economy. This column uses household consumption patterns to examine the distributional impact of the devaluation of the peso during Mexico’s ‘Tequila Crisis’. Cost of living increases are found to be 1.25 to 1.6 times higher for the poor compared to the rich. In the interests of equity, exchange rate policy should take account of such distributional impacts.

The dramatic devaluations of the Russian ruble and Ukrainian hryvnia in 2014, and the instantaneous 20% appreciation of the Swiss franc against the major currencies on 15 January 2015, are only the latest reminders that large exchange rate swings remain a prominent and recurring feature of the world economy. Exchange rates are frequent targets for policies (such as pegs or interventions), and it is important to understand their impact. There is a substantial literature on macro adjustment following large changes in real exchange rates. This literature studies the pass-through to domestic prices, as well as the response of macro aggregates such as imports, exports, and gross domestic product (see Burstein and Gopinath 2015 for a recent survey).

However, there is very little research on whether exchange rate changes affect different types of households differently. This is potentially a first-order phenomenon. A large change in the exchange rate leads to changes in relative prices inside the economy (even though the pass-through to domestic prices is far from complete). It is well known that households at different income levels consume very different baskets of goods. A change in relative prices following a large exchange rate change will thus affect households differentially, implying that exchange rate changes have a distributional as well as an aggregate welfare impact.

In a recent paper we evaluate the distributional impacts of large exchange rate changes using micro data on consumer expenditures and prices (Cravino and Levchenko 2015). We use primarily the devaluation of the Mexican peso during the 1994 Tequila Crisis as the ‘laboratory’ to study these effects. For this episode, we observe monthly price quotes on about 30,000 unique product-outlet level prices that the Bank of Mexico uses to construct the consumer price index, and household-level consumer expenditure surveys both immediately before and after the crisis.