Two Sets of Twins? an Exploration of Domestic Saving-Investment Imbalances

Article excerpt

The United States is on course to increase its net external liabilities to around 40 percent of GDP within the next few years-an unprecedented level of external debt for a large industrial country. This trend is likely to continue to put pressure on the U.S. dollar, particularly because the current account deficit increasingly reflects low saving rather than high investment.

--International Monetary Fund

The twin deficits are back. Their return from the 1980s and early 1990s is now unaccompanied by even a hint that curing the domestic fiscal deficit might well cure the trade deficit, too. The simultaneous expansion of the fiscal surpluses of the late 1990s with the trade deficit dashed that hope. While there is a significant amount of dissonant opinion, a large group of economists is persuaded that the nub of the chronic trade imbalance is the low U.S. saving rate. According to this view, domestic saving has been inadequate to fund current levels of investment. Consequently, offshore capital has moved in to fill the gap, keeping the dollar from falling to levels that would correct the ongoing trade imbalance.

The saving-investment imbalance explanation of the trade deficits can best be understood as a "default" explanation dependent on the failure of traditional trade theory to explain why exchange rate adjustments have been unable to correct the chronic trade imbalance. This is not to say that the textbook story is without a measure of explanatory power. While certain bilateral balances, most particularly the Japanese-USA balance, show positively perverse relationships between the exchange rate and the trade balance, the real weighted exchange rate of the dollar moves in rough inverse correspondence with trade imbalances expressed as a percentage of gross domestic production. (1) It is simply that such adjustments have not been sufficiently powerful. Paul Krugman has given us a partial explanation in suggesting that a lower dollar makes foreign acquisitions of U.S. corporations appear cheap, encouraging a continual inflow of foreign capital (1989, 32), and, indeed, recent cross-border merger and acquisition activity seems to substantiate the claim. (2) But such an explanation begs the question of why trade adjustments are not potent enough to offset such movements of capital. If textbook theory operated unimpeded, the dollar would fall sufficiently to expand exports and restrict imports so as to close the trade deficit and choke off the concomitant inflow of capital. The expansion of exports would generate sufficient domestic income to allow the expansion of sufficient domestic saving so as to equilibrate saving and investment at that higher level of income.

The case for the saving-investment imbalance thesis would be stronger if the dollar had exhibited a strong upper trend corresponding to the "pull" of the U.S. gap between saving and investment. No such trend is exhibited-quite the contrary. For the bulk of the post-Bretton Woods period the dollar has tended downward and, in certain outlier cases, particularly that of Japan, the decline has been spectacular. The proponents of the saving-investment thesis have, of course, recognized this apparent contradiction and have added addenda to their thesis, such as a relative decline of U.S. competitiveness (Krugman 1990, 1994, 118) or lags in the adjustment process relating to the trade balance (Lawrence 1990a, 85; Krugman 1989, 32).

This paper accomplishes two tasks. The first task is to briefly summarize some of the more important literature surrounding the institutional and historical aspects of U.S. multinational trade that have escaped the "clean room" of textbook economics. (3) The intractability of the trade balance after the fall of the dollar in 1985 gave rise to an extensive discussion starting in the late 1980s and early 1990s that was directed at explaining the long-term sluggish response of trade balances to exchange rate adjustments that could not be explained by short-term J-curve effects. …