'Locomotive' Effects
Off the rails?

In the last ten years, policy-makers have periodically called for expansionary action to be taken by other countries, rather than by their own. These calls seem to rest on a belief in the efficacy of the locomotive effect, the idea that expansion by a single powerful economy (the 'locomotive') can stimulate growth throughout the global economy. Growth in the locomotive country, it is argued, will be transmitted to other countries through imports and exports. In Discussion Paper No. 166, Research Fellow Jorge de Macedo examines the validity of the locomotive argument. He concludes that the 'Keynesian' models on which it is based tend to be static and to ignore the intertemporal nature of decision-making; as a result they overstate the 'pull' of the locomotive. He finds that for countries with high levels of external indebtedness, intertemporal effects may even reverse the pull.

De Macedo analyses the transmission of income shocks under flexible exchange rates in a two-country model which explicitly recognizes the existence of such intertemporal links. In the model, the counterpart to the current account balance in each period is a change in foreign assets held. This gives rise to an intertemporal budget constraint for each country, in which the discounted value of present and expected future trade surpluses must offset the current stock of foreign assets. The flexible exchange rate adjusts so as to influence the present and expected balances of trade in order that the sum of their discounted values equals the initial stock of foreign assets. A change in foreign income which causes one of the trade balances to change will bring about an exchange rate adjustment sufficient to restore intertemporal balance.

The transmission of a positive income shock in the 'foreign' country therefore depends on the interaction of income (or locomotive) and exchange rate effects. The first effect is a rise in foreign income which, at an unchanged exchange rate, gives rise to higher foreign imports and home country exports. The resulting improvement in the expected trade balance and income of the home country is the locomotive effect. De Macedo argues that this analysis does not, however, take account of the exchange rate changes necessary to maintain intertemporal balance. Current and expected exchange rates must appreciate in order that expected future trade balances for the home country worsen, offsetting the improvement in the current period. The size of this appreciation will be directly related to the initial level of external indebtedness, which has to be serviced by future exports. If the home country is close to its debt ceiling, therefore, the required exchange rate appreciation may be so large that output in domestic currency falls and the usual 'locomotive' effect will be reversed.

De Macedo simulates the model to show the relative importance of these transmission channels. While the adjusted locomotive channel remains important, the other channels cannot be neglected. He concludes that under flexible exchange rates a cyclical upturn abroad might be less of a locomotive than is commonly assumed, especially for large debtor nations.


'Locomotive' and Other Channels of Transmission under Flexible Exchange Rates
Jorge Braga de Macedo

Discussion Paper No. 166, March 1987 (IM)