Editors' Summary

In macroeconomics, the joke goes, the important questions remain the same, but the answers keep changing. International economists may not be as fickle, but as national economies become more integrated, long-standing policy issues often take on new dimensions and occasionally lead to novel conclusions. The 2001 edition of the Brookings Trade Forum revisits six such issues using new theoretical frameworks as well as new, expanded data sets.

The first two papers examine topics in international macroeconomics and finance. The first of these addresses what factors determine central bank holdings of reserves, and why such holdings have increased even as countries have moved toward more flexible exchange regimes. The second explores whether there is still a role for countercyclical monetary policy--especially exchange rate depreciation--for emerging markets with high capital mobility.

The four remaining papers address topics in international trade. Does the existence of visible international conventions affect the probability that individual countries will adopt and implement labor standards? Why are some people more protectionist than others, and how well are these attitudes explained by the types of economic factors emphasized in standard trade models? How important are the remaining barriers to cross-border trade, and would removal of such barriers significantly improve welfare? Finally, what determines a firm's decision to invest abroad? In particular, how important are market access and wage differentials relative to other factors?

The Brookings Trade Forum held its fourth annual conference in Washington on May 10-11, 2001. This issue of the Trade Forum contains the six papers and discussion from that conference.

IN THE FIRST PAPER, Robert Flood and Nancy Marion examine why countries held so many international reserves during the 1990s. At the end of 1999, global reserve holdings (excluding gold) were almost double what they had been at the end of 1960 and about 20 percent higher than they had been at the start of the 1990s. Reserve holdings have also trended upward when measured as a share of global income. A primary use of international reserves is for countries to stabilize the exchange value of their currencies. However, by standard measures, most countries pay less attention to currency stabilization now than in the recent past. They certainly pay far less attention to currencies' exchange values than they did in the Bretton Woods era. As pointed out by discussants and other participants, this topic has received surprisingly little recent attention in the economics literature.

Flood and Marion begin by carefully documenting what has happened to reserve holdings among a broad sample of countries in the post-World War II period. Their paper then turns to more formal statistical analyses in an attempt to explain the observed patterns. The authors first replicate and update early econometric estimates of countries' reserve-holding behavior based on a "buffer-stock" model. This early work developed the insight that countries should tend to hold more reserves when reserve levels are more volatile. However, early approaches for measuring this volatility are likely to be biased, as explained in some detail in the paper. Therefore, Flood and Marion extend the methods and data used in previous econometric work. By building on more recent models in which reserve movements respond to government policies and countries may be subject to speculative attacks on reserve stocks, they construct better measures of reserve volatility that can be used in the estimation. Finally, after estimating models constrained tightly by suggested economic theory, the authors report some exploratory work that considers the importance of other factors in explaining recent reserve holdings.

Overall, the results reported in the paper are mixed. The replication and data extension of previous work indicates that the buffer-stock model works about as well using the extended data set as it did for 1970s data. Reserve holdings increase with greater reserve volatility but do not respond significantly to their opportunity cost. Using their improved measure of volatility the authors show that their results are very robust. Their new volatility measure is always a highly statistically significant explanatory variable for reserve holdings.

However...