We have seen thus far that exports do not exhibit consistent and systematic long swings from 1820 to 1913, and for that reason we treat them as unsystematic relative to our long-swing framework. Either exports are too dependent on variables outside the indigenous system (i.e., world prices, foreign demand and supply) or the effects of the domestic long swing upon export components are much too varied to produce a systematic pattern in aggregate exports directly or inversely related to domestic activity. This does not exclude the possibility, of course, that exports played an important role before the Civil War, if not by initiating the early Kuznets cycle at least by reflecting and supporting the domestic long swing.
Conceptually, the problem seems much simpler in the case of imports. If domestic price movements are correlated with realincome movements over the long swing, contradictory pressures will be exerted upon exports; a priori, this seems unlikely in the case of aggregate imports. On the contrary, not only will real growth stimulate import demand under normal conditions (i.e., where the income effect exceeds the rate of import substitution), but also price inflation will add to this increased demand for for-