The Houthakker-Magee (HM) effect is one of those empirical regularities, like Okun's Law, that suggests that Kaldor was correct when he talked about stylized facts, and that indicates that economics might actually aspire to be a more serious scientific endeavor (in spite of economists, of course). The HM effect says that the income elasticity of imports in the US tends to be about double the size of the income elasticity of exports, and as a result the US has a tendency to run trade deficits if it grows at the same pace than the rest of the world.
Several more recent estimates of the HM (the original Houthakker and Magee paper is from 1969*; subscription required) confirm the persistence of the phenomenon. Peter Hooper and Jaime Marquez are responsible for more than one of those more recent estimates (see, for example, here with Karen Johnson), which also suggest that income elasticities are considerably larger than price elasticities. In particular they note that, at least for G7 countries, imports are not responsive to changes in the real exchange rate. They estimate that US export and import income elasticities are 0.8 and 1.8 and that everything else constant, and, if the US grew at the same pace than the rest of the world, a depreciation of 2.8% in real terms per year would be necessary to maintain the trade balance.
Mind you, the question of whether the US needs the depreciation (at the point they wrote the paper, actual depreciation had been more like 1.1% per year) is a different question. The role of the dollar allows the US to run deficits that other countries could not, so the external constraint does not play a role in this case. But that was discussed before (here).
* Note that Houthakker and Magee cite in their paper the previous work by Neisser and Modigliani done at the New School, which was, as far as I know, the first estimation of the US trade elasticities.
Several more recent estimates of the HM (the original Houthakker and Magee paper is from 1969*; subscription required) confirm the persistence of the phenomenon. Peter Hooper and Jaime Marquez are responsible for more than one of those more recent estimates (see, for example, here with Karen Johnson), which also suggest that income elasticities are considerably larger than price elasticities. In particular they note that, at least for G7 countries, imports are not responsive to changes in the real exchange rate. They estimate that US export and import income elasticities are 0.8 and 1.8 and that everything else constant, and, if the US grew at the same pace than the rest of the world, a depreciation of 2.8% in real terms per year would be necessary to maintain the trade balance.
Mind you, the question of whether the US needs the depreciation (at the point they wrote the paper, actual depreciation had been more like 1.1% per year) is a different question. The role of the dollar allows the US to run deficits that other countries could not, so the external constraint does not play a role in this case. But that was discussed before (here).
* Note that Houthakker and Magee cite in their paper the previous work by Neisser and Modigliani done at the New School, which was, as far as I know, the first estimation of the US trade elasticities.
