The brookings panel on economic activity held its eighty-seventh conference in Washington, D.C., on April 2 and 3, 2009. The conference occurred barely six months after the collapse of the investment bank Lehman Brothers, an event often used to date the transition from a largely conventional cyclical downturn, characterized by a strained financial system and mild recession, to a full-blown financial and economic crisis. In keeping with the Brookings Papers' tradition of providing timely analysis of current economic events, three of the papers in this volume address the role of various factors in the initial downturn and the ensuing crisis, including the response of policymakers, the behavior of bond markets, and the role played by the oil market. The two remaining papers examine the impact of tax cuts on government spending, and the role of corruption in undermining popular support for market-oriented policies.
In the first paper in this issue, Phillip Swagel provides an insider's account of the policy debates as they unfolded in real time during his tenure as assistant secretary of the Treasury for economic policy in the last two years of the Bush administration. Swagel's account is both a blow-by-blow history of the policy response to the crisis and a lesson in economic realpolitik. He documents that the Treasury under Secretary Henry Paulson was quite aware of the fragility of the financial system as early as 2006, and indeed, interagency work was well under way to develop a strategy for dealing with a crisis should one occur.
Swagel's narrative provides insight into the constraints on the policy process that were not immediately apparent to many commentators. In particular, policymakers at the Treasury and at the Federal Reserve were unable to pursue a number of useful policy proposals simply because they and other government agencies lacked appropriate legal authority. Although Congress could have granted that authority, this raised the even [End Page vii] larger concern of political constraints, which were particularly important in the context of a profound lack of trust between the executive and legislative branches. A recurrent theme of the paper is the sheer difficulty of getting Congress to respond under anything less than crisis conditions—and possibly even then—which delayed and diluted the eventual policy response. Swagel also highlights a third, more practical constraint: time. Policymakers had to make decisions rapidly, often with too little information, as financial markets collapsed around them. He reserves some constructive criticism for the role played by many academic macroeconomists throughout the ensuing public debate: their editorializing, in his view, appeared largely uninformed by the various constraints, rendering their advocacy often unhelpful and occasionally even counterproductive. At the same time, however, he faults the Treasury for doing a poor job of making its case to a skeptical public.
In the second paper, John Campbell, Robert Shiller, and Luis Viceira present a thorough accounting of what has been learned from the first quarter-century of experience with inflation-indexed bonds in the United Kingdom and the first decade of experience in the United States. Yields on these bonds indicate a substantial and puzzling decline in long-term real interest rates from the 1990s through 2008. The volatility of these real rates was likewise unexpected, given that a key determinant, the marginal product of capital, can reasonably be presumed to be stable over time. Over the same period, movements in the prices of inflation-indexed bonds have come to be negatively correlated with movements in stock prices. The authors also find that seemingly very similar bonds can bear surprisingly different yields, with real U.S. and U.K. yields diverging at times by over 2 percentage points.
Having documented these facts, the authors set out to explain them. They begin with the expectations theory of the term structure—the view that long-term real yields reflect current and expected future short-term real interest rates. As expected short-term real rates vary, so too does this long-run expectation. Using a simple econometric model to proxy for expectations about current and future short-term rates, the authors succeed in replicating some of the observed changes in long-term inflation-indexed bond yields. Even so, the actual yields are higher and more volatile than suggested by these expectations. This leads the authors to explore whether the yields include a positive risk premium, and if so, how it has varied through time. However, the fact that inflation-indexed bond prices are negatively correlated with equity prices suggests that the risk premium should be negative, reflecting the value of these bonds in portfolio diversification. [End Page viii] Moreover, changes over time in this correlation can explain very little of the changes over time in real bond yields.
Finally, the authors pay special attention both to the high yields on inflation-indexed bonds in the years following their introduction, and to the extraordinary volatility in these yields during the financial crisis in 2008. They conclude that institutional factors played an important role in both phenomena. Such "technical" factors are typically invoked as a euphemism for changes that cannot be otherwise explained, but the authors probe more deeply, finding that the current episode is "highly abnormal" and likely due to illiquidity, as financial institutions were forced to unwind large positions quickly and few of the usual buyers were able to absorb these large shifts. Despite the apparent pricing anomalies that they document, however, the authors conclude that inflation-indexed bonds provide both a useful investment instrument for many investors and a valuable financing tool for governments.
In the third paper, Christina Romer and David Romer examine the "starve the beast" hypothesis. (In the interest of full disclosure, we note that this paper was commissioned before David Romer accepted the editorship of this journal, and that the editorial duties were handled by his co-editor.) This hypothesis, most closely identified with President Ronald Reagan and his advisers, holds that tax cuts today impel future reductions in government spending, and hence in the size of government. Some previous studies have found evidence tending to support the hypothesis, but Romer and Romer point to two inherent problems in testing it. One is that the observed correlation between tax cuts and changes in spending might reflect reverse causality, with changes (typically increases) in government spending priorities causing changes in taxation. The other is that some third factor, such as a slowing economy, might affect both taxes and spending, producing a spurious association between the two.
To try to isolate the effect of tax changes on government spending, the authors rely on a narrative approach, poring over presidential speeches, congressional reports, and other documents to identify those legislated tax changes not tied to either spending changes or the state of the economy. This yields a set of tax changes that, they argue, can appropriately be used to test the starve-the-beast hypothesis. The most important of these are the 1948 tax cut that passed despite President Harry Truman's veto, the Kennedy Johnson tax cut in the mid-1960s, the Reagan tax cut in the early 1980s, and two tax cuts passed during President George W. Bush's first term.
In an exhaustive analysis of this more restricted set of tax changes, Romer and Romer find remarkably little evidence in favor of the starve-the-beast [End Page ix] hypothesis. The tax cuts they identify are not followed by any systematic decrease in government spending relative to its previously expected path; indeed, there is some evidence of a tendency for spending to increase. Moreover, based on the documentary evidence, policymakers in the major episodes appeared largely unconcerned about the sufficiency of revenue when making their spending decisions. Instead, the subsequent budgetary adjustments in these episodes generally consisted of a combination of legislated tax increases and nonlegislated increases in revenue. The popular view that tax cuts are "sticky" and not easily undone, and thus, that they create strong pressures for reductions in spending, appears false.
The fourth paper, by James Hamilton, returns to a theme to which he has already made major contributions: the macroeconomic consequences of oil price shocks. The price of oil has recently been spectacularly volatile, rising from $60 a barrel in mid-2007 to a high of $145 a barrel in mid-2008 before collapsing to $30 a barrel by the end of 2008. Hamilton notes that this oil shock differed sharply from previous disruptions. Whereas those could typically be traced to geopolitical developments, this time the rise in oil prices reflected the working of the price mechanism to reconcile stagnating worldwide production with rising demand, particularly from China. Because the price elasticity of demand for oil is low, it took a large price rise to bring the quantity demanded back into line with the quantity supplied. In Hamilton's account, perhaps the most important "shock" during this period was to the received ideas of market participants, who learned that short-run demand is even more price inelastic than they had thought. This shock helps explain both why oil production did not rise in response to the growing demand and why oil inventories declined. The popular competing story—that speculators bid up the price of oil futures, driving up the spot price—faces the key difficulty that it suggests, counter-factually, that inventories should have been accumulating. Hamilton argues further that the subsequent collapse in oil prices in 2008 was too large to be explained by the global recession. Instead, it likely reflects forces similar to those that accounted for its rise, but in the opposite direction, as market players underestimated the medium-run price elasticity of oil demand.
Hamilton shows that the oil price rise had quantitatively important effects in reducing both consumption spending—particularly on motor vehicles, and especially on domestically produced SUVs—and consumer sentiment in the United States. Marshalling an array of macroeconometric models, he shows that the oil shock explains a large part of the weak macroeconomic performance in late 2007 and much of 2008, and indeed [End Page x] that in the absence of the shock, this period would have been one of slow growth rather than outright recession.
The final paper of this issue, by Rafael Di Tella and Robert MacCulloch, begins by documenting an important and previously overlooked stylized fact, namely, that capitalism is surprisingly uncommon around the world. Outside the major industrialized nations, heavy business regulation, leftist rhetoric, and pro-intervention beliefs are the norm. As the authors stress, this pattern is puzzling, given the strong evidence of the positive effects of market-oriented policies on economic growth and the enormous potential gains for poor countries, in particular, from adopting such policies.
The authors propose an intriguing explanation of this seemingly perverse bias against capitalism in poor countries: that it reflects the response of their citizens to a corrupt business sector. Corruption, the authors argue, is seen in these countries as disproportionately the fault of business leaders who seek favors from government rather than the bureaucrats who grant them. Hence corruption undermines the legitimacy of business, leading voters to favor interventionist and anticapitalist policies as "punishment" against the business owners, who are broadly seen as having benefited unfairly from their favored position in society. The authors document that within a country, those individuals who perceive corruption to be more widespread tend to have more-interventionist beliefs. They also show that sharp increases in corruption within a country lead to a rise in left-wing voting. Anger is also found to be a more prevalent emotion in corrupt countries than elsewhere, and it is lessened by stricter regulation of business.
We conclude this summary by noting the recent changes in the editorship of the Brookings Papers. Although this is the first volume with our names on the cover, the papers were selected by the previous editorial team of Douglas W. Elmendorf, N. Gregory Mankiw, and Lawrence H. Summers; William G. Gale handled much of the work during the transition to the new team. All four have our thanks for their work on this volume and throughout their tenures. We hope to build upon the outstanding foundation built not only by these editors but also by their predecessors, William C. Brainard, Arthur M. Okun, and George L. Perry. We will strive to ensure that the Brookings Papers continues to be a key source for original, timely, and substantive analysis of a wide array of pressing economic issues. [End Page xi]