The excellent Douglas Irwin has a new NBER paper on that question, here is one excerpt:
Hayek (1937, 64) leveled three main criticisms against flexible exchange rates, all of which were frequently repeated during this period. First, flexible exchange rates would give rise to speculative capital flows that would be destabilizing; specifically, capital movements would reinforce exchange rate shifts arising from payments imbalances, thereby magnifying volatility and “turn what originally might have been a minor inconvenience into a major disturbance.” Second, flexible exchange rates would lead to competitive depreciations, the flexible rate counterpart to competitive devaluations, which would encourage a return to mercantilism and an increase in trade barriers. “Without stability of exchange rates it is vain to hope for any reduction of trade barriers,” he concluded (1937, 74n). Third, exchange rate instability would create risks that would discourage international trade and deter long-term foreign investment.
Frank Graham and Charles Whittlesey, both at Princeton, were among the few American economists who favored complete floating rates and monetary independence. Now what might account for such a difference in opinion?:
1. They hadn’t yet learned that fixed rate systems just weren’t politically stable, but we now know this with the benefit of hindsight, including the failures of Bretton Woods and a new understanding that competitive devaluations don’t have to be so disastrous.
2. They were good economists, and we are plain, simple idiots.
3. Heavy-duty manufacturing exports, with only a few major exporting countries, and a lot of FDI potential in the periphery, plus plenty of highly illiquid currencies, actually militated in favor of fixed rather than floating rates at that time.
4. During that period people thought high level of international cooperation were necessary to solve problems, and this stemmed in part from the failures of World War I and later World War II. If you favor “international cooperation” as a general value, you might then also tend to mood affiliate with the notion of fixed exchange rates.
I believe that factors #1-4 all might play a role in the complete explanation here. Am I overlooking something?