DESPITE occasional manifestations of disappointment and distrust, the globalization of economic life is now almost taken for granted. Nowhere has this trend been more pervasive than in global financial markets in the past few decades. Capital flows have surged in volume, in both the developed and the developing world, creating new opportunities for economic benefit and difficult challenges for policymakers. The dust has by no means settled.
It may surprise some readers to learn that the paragraph above describes not only 2004 but also 1904, during the era of globalization that spanned the years 1870 to 1914. The striking parallels between that era and the current era of globalization have been described in many recent studies. These parallels raise a number of questions about the evolution of the global economy in the 19th century, its collapse in 1914, and the rebirth of globalization at the end of the 20th century.
With Maurice Obstfeld (University of California, Berkeley), I have explored these events systematically, which resulted in our new economic history of global capital markets and the attendant political-economy problems. We found that economic policymaking has, throughout, been characterized by a fundamental macroeconomic policy trilemma that all governments face. That is, it is not possible for a government simultaneously to peg the exchange rate, keep an open capital market, and enjoy monetary policy autonomy. In this article, I synthesize our evidence for the pervasiveness of the trilemma and draw lessons for today’s policymakers.
Policymakers in two eras of globalization faced the same “trilemma” of difficult policy trade-offs
Alan M. Taylor
Rise and fall of globalization
In the early 19th century, international finance was, in many respects, a fairly direct descendant of its 17th-century predecessor and was still dominated by London and Amsterdam, Institutionally, the more or less laissez-faire attitudes that had prevailed since 1688 allowed these markets to mature. But the volume of capital remained small and was confined mostly to finance within Europe, and technological progress was slow. However, this was all to change in a matter of a few decades, with the development of a global financial market and its assorted handmaidens—the telegraph and other improvements in transportation and communications; the increasing rate of growth of European settlement; and the arrival, through imposition or imitation, of institutional “modernization.” Of particular importance for the story we tell was the emergence of the classical gold standard as an international monetary regime. The world economy of 1913 was vastly different from the early 19th-century one.
But this economy imploded under the strain of the two world wars and the Great Depression, as well as under the political-economy tensions that accompanied this era of unprecedented upheaval. By the mid-1930s, the free flow of goods, people, and capital was almost at a standstill. The better part of the 20th century—at least since 1929 and perhaps since 1913—is a tale of radical experimentation in political economy and monetary policy that naysayers, beginning with economic historian Karl Polanyi, predicted would doom economic integration for good. For many decades, they appeared to be right.
Enter John Maynard Keynes and Harry Dexter White, architects of the postwar economic order known as the Bretton Woods system, with the IMF as one of its foundations. The IMF embodied a new macroeconomic paradigm, with currency pegs and capital controls as cornerstones. Because the role of free capital flows in the crises of the 1930s had come under suspicion, the IMF espoused capital controls. Similarly, floating rates were associated with speculation and instability and, hence, the disruption of trade. These fears motivated the choice of fixed (albeit “adjustable”) parities. This blueprint makes it clear why, even after 1945, global capital markets took so long to recover.
Only after the Bretton Woods system unraveled under the strain of balance of payments pressures in the late 1960s and early 1970s did a new order begin to coalesce, and, even then, it did so only fitfully. The unraveling began when it became clear that for trade to be sustained at a volume that would deliver meaningful economic benefits, large payments transactions would need to be allowed. And, in response to political-economy tensions, the currency peg was being adjusted with some regularity. Not only was this a recipe for crisis in the short run, but it would lead, in due course, to questions about the overall design of the global financial architecture. Furthermore, the potential gains from financial openness were becoming apparent as developing countries, for the most part closed to flows from the developed economies, labored under financing gaps. Ultimately, the vision of Keynes and White of a world that could be kept safe for trade by constraining private finance turned out to be an illusion.
Skeptics, however, warned that the transition to floating rates would loosen the reins on finance at the price of disrupting trade and perhaps even financial flows themselves. We now know that trade has flourished ever since (see Chart 1), as has finance, with both now flowing at volumes that, by some yardsticks, exceed the peak reached in 1913.
Chart 1Capital overflow
The growth of the global capital market in both eras of financial market integration was impressive.
Source: Obstfeld and Taylor, 2004.
Note: The chart shows the ratio of the stocks of international investments (measured by gross assets) to gross domestic product. The sample comprises the major capital exporters and other countries that enter the sample over time. For details of the changing sample, see Obstfeld and Taylor; 2004.
How have governments’ policy decisions influenced the ebb and flow of capital over time? The trilemma serves as an organizing principle for a discussion of the monetary history of global capital markets. It is a way of describing governments’ choice from three policy goals: pegging the exchange rate, keeping the capital market open, or conducting an activist monetary policy. The trilemma arises because a government can achieve only two of those policy goals at any one time. For example, it can achieve exchange stability and an open capital market by adopting a permanently fixed exchange rate, but must give up monetary independence. If a government opts for monetary independence and an open capital market, it can float the exchange rate but cannot achieve exchange stability. Finally, if a government chooses exchange stability and monetary independence, it abandons the goal of capital market integration.
The trilemma, then, should have major implications for monetary policy. But is it an empirically important phenomenon? This has been a difficult question for economists to answer. With Jay C. Shambaugh (Dartmouth College) and Obstfeld, I tried a new tack: exploring the comovement of domestic and foreign interest rates. Over 130 years, we classified the degree of comovement according to the exchange rate regime in place (fixed or floating) and the presence or absence of capital controls. A higher degree of comovement implies that a change in foreign interest rates translates quickly into a similar change in domestic interest rates and, hence, implies the absence of monetary autonomy of domestic authorities. Monetary autonomy was constrained, for example, when exchange rates were fixed and capital accounts were open (under the gold standard), but prevailed when exchange rate regimes were fixed but capital accounts were relatively closed (under the Bretton Woods system).
A good place to begin the review of the evidence is the period of the gold standard (roughly 1870–1913), which remains, in some ways, the benchmark for a rigid exchange rate regime. However, even the gold standard allowed minor room for maneuver because of the existence of gold points—small bands in exchange rates that arise on account of transaction costs. In such a setting, referred to by Paul Krugman (Princeton University) as a target zone, Lars Svensson (also of Princeton) has formally derived the term structure of interest rates and shown that large short-term interest rate differentials can emerge. Thus, domestic and foreign interest rates may not move in lock-step. So, in this model world, imperfect pass-through of foreign to domestic interest rates clearly cannot be read as an indication of policy autonomy; it may represent limited autonomy as constrained by the target zone. The real-world problem is therefore ultimately an empirical question of how tight those constraints are.
Not surprisingly, under the gold standard, the naive interest parity result of complete pass-through from foreign to domestic interest rates (a coefficient of one) does not hold, even with the quite narrow bands we impose to approximate the gold points. We find a coefficient of 0.6, which is statistically different from both zero and one. Simulations of a target-zone model show that 0.6 is the coefficient to be expected with an exchange rate band of about 1 percent (between the gold points) and essentially perfect capital mobility.
In contrast, during periods of floating exchange rates, this coefficient was found to be closer to zero—which means that, theoretically, domestic policymakers recovered considerable room for maneuver when they abandoned the peg. For example, in the (not infrequent) floating episodes during 1870–1913, countries operated with considerable monetary policy autonomy; the pass-through coefficient was close to zero, which was comparable to that predicted by the simulation for parameters during that period.
When the same analysis was carried out on the Bretton Woods peg, no pass-through from foreign to domestic interest rates was observed. Thus, it may be inferred that when the Bretton Woods architects made it their mission to limit capital mobility so as to free domestic monetary policy, they succeeded admirably. When more refined dynamic models are used, some developed countries later in the Bretton Woods period are found to be exceptions to the rule, but the general result stands.
Since the collapse of the Bretton Woods system, however, the world has begun to resemble the classical gold standard era once again. Overall, pegging may be less frequent than it was 100 years ago, but when countries do elect to peg in the prevailing environment of increased capital mobility, the trilemma hampers their ability to conduct monetary policy as much as ever. The pass-through is about 0.5 and is statistically indistinguishable from the 0.6 recorded under the classical gold standard. Contemporary floats, in contrast, show a coefficient that is much smaller, as expected. The pass-through coefficients under floats are not as close to zero as during the late 19th and early 20th centuries, perhaps indicating some policy convergence of other nominal anchors (for example, inflation targeting) that might impose common monetary policy goals even on potentially autonomous policymakers in different countries. But the modern float and peg coefficients are still significantly different from each other, so we can assert that the basic trade-off embodied in the trilemma holds true.
“The unraveling of Argentina in 2001 offers a classic example of creeping monetary policy autonomy … leading to the collapse of a peg under open capital markets.”
We conclude that, in general, over a wide range of historical experience covering more than 100 years, floats have permitted more interest rate independence than pegs, except when pegs were combined with pervasive capital controls, as under the Bretton Woods regime. Our ongoing research on the tumultuous interwar period, when classification of regimes was more difficult and data problems more acute, only reinforces this conclusion. This body of research supplies quite convincing evidence that the trilemma has endured as an important constraint on what is feasible for macroeconomic policymakers.
Lessons for policymakers
Obeying the dictates of the trilemma can, of course, be politically costly. All too often, politics has trumped economics, and past experience illustrates the dangers that can arise. The peril, of course, is that policymaking will veer into inconsistency, markets will interpret countries’ intentions as so much nonsense, and a crisis will ensue. This is the recurring nightmare of emerging market economies, whose crises are generally much more painful and prolonged than those in developed countries. Collateral damage to the financial system is often severe, and the political-economy ramifications—such as government instability and lost reputation—tend to be harsher. Although the developed countries appear to have overcome historical problems in managing global capital—a severe challenge in the 1920s and 1930s—the developing economies still struggle. For example, the unraveling of Argentina in 2001 offers a classic example of creeping monetary policy autonomy (in particular with respect to the fractional reserve banking system and through the issue of government-guaranteed quasi monies through the state and federal treasuries) leading to the collapse of a peg under open capital markets. The exchange rate floated far away, and a semblance of stability was restored to the exchange only when capital controls were imposed. Policymakers have to remember the constraints of the trilemma; when these are ignored, as they were in several recent crisis episodes, the costs can be grave.
Does this observation imply that the global capital market might benefit only the developed economies? This raises the issue of the very different function of the market for rich and poor countries. A hundred years ago, a significant fraction of the financial flows emanated from rich countries (especially Great Britain) and headed for underdeveloped regions, both within and beyond the Empire, The flows carried development finance from capital-abundant regions to capital-scare regions and were intended to facilitate long-run growth. These days, very little capital flows to poor regions. Most global flows are diversification finance—almost offsetting gross flows from one rich country to another—and are intended largely to reduce risk through the fine-tuning of portfolios.
Today’s global capital market does not deliver sufficient capital to help the neediest countries in their long-run development (see Chart 2). But is this market failure or government failure? Recent research suggests that many poor countries are, in fact, not very far from their long-run “steady state”—that is, the long-run equilibrium where the growth of capital per worker and output per worker level off given a country’s preferences for saving and technology for production. In most models, a country’s long-run destiny is determined by productivity and patience, not by the relative availability of local versus foreign capital. For technological determinists, if today’s poor countries have indeed reached their steady states, this is not so much a flattering statement about their domestic saving capacity as a devastating comment on their social, political, institutional, educational, technological, and ecological barriers to productivity advance.
Chart 2Who benefits?
Foreign capital used to flow to poor countries, but now flows mostly to rich countries.
Source: Obstfeld and Taylor, 2004.
Does this render discussion of the benefits of a financially integrated world order moot? Suppose that the productivity doctors get to work in developing countries, which start to catch up. Where would this leave them? Far from their steady states. Capital market integration is of little use to people unfortunate enough to live in a dysfunctional system and, by extension, in a desperately poor economy. It must not be sold to them for the panacea it is not. But clearly, for those countries that can get on the escalator of modern economic growth, the external financing margin becomes much more important because they have something worth financing. Our theoretical simulations develop this intuition and empirical work confirms it. The data show that institutionally weak countries gain little from financial opening, but those that engage in serious reforms stand to reap large gains.
Echoes of the past
In many ways, the 19th century was a relatively simple context in which to build an international financial architecture for the first era of globalization. Democratic politics were still distant; the powerful sway of financial orthodoxy coupled with elite-led governance speeded the creation of a global capital market; institutions like the gold standard and the extension of Empire made the ride even smoother.
From the interwar period until the past few decades, global capital markets almost shrank from sight. But they have proved more resilient than many observers had expected. Still, the rebirth of globalization should not cause us complacently to consider that the dust has settled once again. Echoes of past crises and policy failures still reverberate, and all countries should heed the message to learn from, rather than repeat, past mistakes. As governments begin to comprehend the limits to autonomy—in every sense—and the gains from integration, the whole world stands to benefit.
Alan M. Taylor is a Professor of Economics and Chancellor’s Fellow at the University of California, Davis; a Research Associate of the National Bureau of Economic Research; and a Research Fellow of the Centre for Economic Policy Research.
This article is based on the book Taylor coauthored with Maurice Obstfeld, Global Capital Markets: Integration, Crisis, and Growth (Cambridge: Cambridge University Press, 2004).
INTERNATIONAL MONETARY FUND
WASHINGTON D.C. 20431
HELP US IMPROVE FINANCE & DEVELOPMENT (F&D)
We need your feedback to help us find ways to better meet your needs. Often, people who provide feedback tend either to know us very well or else to have very strong views about something they have read. But we want to hear from a large and truly representative sample of our readers their opinions of Finance & Development. Armed with this information, we will be in a better position to effect improvements – to make it a better, more enjoyable, and more useful magazine.
Therefore, please help us to help you by taking the 15 minutes necessary to fill in and send back this questionnaire. Answer all the questions you can as candidly as possible. If a question doesn’t seem to apply to you, skip it and move on to the next question. The more of you who answer and send back the questionnaire, the more representative and reliable will be the results.
You will note that we have not asked for your name and address, only for the name of your country of residence. The questionnaires will be analyzed for us by a market research company and data-processing bureau – Digitab – in the UK, hence the UK return address. All responses will be treated with complete confidentiality.
Your views may truly make a difference – so do express them! Thank you for your patience.
Finance & Development