SOME EXPERIENCES UNDER PAPER MONEY 399 ditions — always the proximate cause of movements in inter- national trade — checked exports during the first stage, imports during the second. The mechanism was different from that of gold exchange prices based on the gold standard, but the sub- stantive outcome was the same. Not everything, however, is to be explained in this way. We must be on our guard here, as in all statistical and historical inquiry, against over-simplification in interpreting phenomena that are confused and complicated, often remaining obscure after the most painstaking examination. In the present case, for exam- ple, a further factor goes to explain the heavy imports into the United States during the years 1869-73. This was the direct purchase of British goods by the American borrowers. The way in which the balance of trade is affected by the direct linking of loans with expenditures in the lending country has been analyzed else- where.! These effects are the same under paper conditions as under gold conditions. So far as the proceeds of the loans were applied in part to the immediate purchase of goods in Great Britain, and so far as these purchases were additional to what would have been bought by Americans in any case, there was at once a movement of exports from Great Britain to the United States and an increase of imports into the United States. Operations of this kind could have no effect on the foreign exchange market or the gold premium in the United States, or the prices of goods in the United States. Such transactions call for no such recondite explanation as has been undertaken in the preceding paragraphs. None the less, there remain transactions to which the simpler explanation cannot apply. By no means all the proceeds of British loans were used in direct purchases. A large part — the larger part, one would guess, for these particular operations — was cashed in by the borrowers, so to speak. Funds were wanted by them for use in the United States, and remittances to the United States were called for. And then we do need to resort to the more recondite explanation, and to search for evidence which will cor- roborate it. Chapter 12, p. 127