Sections

Commentary

Editors’ Summary of the Spring 2008 Brookings Papers on Economic Activity

Douglas Elmendorf,
Douglas Elmendorf Harvard University Distinguished Service Professor - Harvard Kennedy School
N. Gregory Mankiw, and
N. Gregory Mankiw Robert M. Beren Professor of Economics - Harvard University
Lawrence H. Summers
Lawrence H. Summers Charles W. Eliot University Professor and President Emeritus - Harvard University

April 10, 2008

The papers discussed below are available as free downloads at the Brookings Papers on Economic Activity archive.

The Brookings Panel on Economic Activity held its eighty-fifth conference in Washington, D.C., on April 10 and 11, 2008. For this conference—our first as editors of the journal—we selected papers on a range of topics central to macroeconomics and economic policy: the rela­tionship between economic growth and people’s subjective well-being; the effect of international trade on wages; the appropriate role of policies designed to strengthen local economies; macroeconomic crises and asset pricing; the impact of political constraints on the success of economic policy reform; and the behavior of financial markets during the approach of world wars. This issue of the Brookings Papers on Economic Activity presents the six papers from the conference, comments by the formal discussants, and synopses of the discussions of the papers by conference participants.

In the first paper, Betsey Stevenson and Justin Wolfers examine the effect of income on subjective well-being. More than thirty years ago, Richard Easterlin compiled empirical evidence suggesting that an increase in a country’s average income did not increase the average level of happi­ness among its citizens. This “Easterlin paradox” cast doubt on the desir­ability of public policy focused on increasing material standards of living. However, Stevenson and Wolfers argue that the data available for assess­ing the link between income and happiness were quite limited for many years, and that the apparent lack of a robust link led some observers to “confound the absence of evidence of such a link with evidence of its absence.”

Today, data on income and self-reported satisfaction are available for a broad array of countries at all levels of development over a number of decades. Using an extensive collection of these new datasets, Stevenson and Wolfers estimate that income has a significant positive effect on subjective well-being. Moreover, the magnitude of the estimated income­ happiness gradient is similar across countries, among people of different income levels within countries at a point in time, and within countries as their average income changes over time. In particular, the authors find no evidence of a satiation point beyond which higher income does not raise well-being.

These results show that absolute levels of income have important ef­fects on well-being. However, the authors acknowledge that relative in­comes may matter as well: Because their estimates of the income-happiness gradient are imprecise, one cannot rule out either that the within-country effect of income is smaller than the cross-country effect (which might be attributable to relative-income effects) or that these two effects are identi­cal (which would speak against a role for relative income). The authors also note one striking exception to the pattern they observe: Americans have experienced virtually no increase in happiness during the past thirty-five years despite ongoing increases in average income.

In the second paper, Paul Krugman returns to a question that he and others studied in the 1990s: How does international trade affect the wages of less skilled workers in the United States? Textbook analysis predicts that increased trade with countries that have large quantities of unskilled labor should reduce the relative price of goods whose production uses unskilled labor intensively, and that this reduction in relative price should reduce the relative wages of less skilled workers in the United States. The key question is whether this effect is quantitatively important, and the answer given by several studies in the 1990s was that U.S. trade with developing countries was having only a modest effect on U.S. income inequality.

Krugman explains, however, that the magnitude and nature of trade have changed considerably since these earlier estimates were made. Im­ports from developing countries have roughly doubled as a share of U.S. GDP since the early 1990s. Moreover, the developing countries that account for most of the recent expansion in U.S. trade have substantially lower wages, relative to U.S. wages, than did the developing countries that were the focus of the previous literature. Therefore, the effect of trade on income inequality is potentially much larger today.

Examining the data, Krugman documents a marked increase in imports of products that are traditionally skill intensive, such as computers and other electronic products, compared with products that are not skill inten­sive, such as apparel. Taken at face value, this pattern would suggest that trade is not tending to widen the U.S. income distribution. But Krugman deems it quite unlikely that the surge in imports from low-wage countries is concentrated in value added by high-skilled workers, and he speculates that falling costs of doing business internationally have enabled develop­ing countries to specialize in low-skill niches within generally high-skill industries. Even so, Krugman concludes that the shifting pattern of inter­national trade has probably been a fairly small factor in the widening of the U.S. income distribution.

In the third paper in this issue, Edward Glaeser and Joshua Gottlieb examine the effect of government policies focused on local economies— what they term “place-making policies.” They begin by presenting evi­dence that regions in the United States are close to a spatial equilibrium, in which higher wages in a geographic area tend to be offset by higher prices and lower amenities. Under these conditions, shifting resources from places with higher average incomes to places with lower average incomes would shift land prices and other prices correspondingly; these shifts would not enhance equity.

Glaeser and Gottlieb then document the concentration of economic activity in densely populated clusters and the strong positive connection between density and productivity. The existence of such “agglomeration economies” creates a potential justification for government intervention. If policymakers could identify the areas where additional density would have the largest effect on productivity, encouraging growth there would be use­ful. Unfortunately, identifying such areas is very difficult; for example, the authors find little evidence that agglomeration economies are more impor­tant for smaller cities than for larger cities, or for more compact cities than for less compact cities. Similarly, concentrations of educated people seem to generate an extra boost to average productivity. These positive spillovers imply that local leaders should aim to attract or train more skilled residents. Yet any effort by national leaders to induce skilled people to move from one place to another would be useful for the country as a whole only if analysts knew where the benefits from additional skills were the highest, and that is not known.

In addition, the authors contend that investment in transportation infra­structure should be geared not toward making certain areas more desirable relative to others, but instead toward generating the largest direct benefit for households where they currently live. They also encourage more build­ing in high-income areas that currently restrict new construction through tight controls on land use.

The fourth paper, by Robert Barro and José Ursúa, examines macro­economic crises around the world since 1870. The authors build on the long-term international data on aggregate output compiled by Angus Mad­dison, both improving on the measurement of output and assembling cor­responding measures of aggregate consumer spending. Barro and Ursúa argue that Maddison’s data, although a valuable resource for economists, contain a number of problematic observations. Some of these problems arise from the assumptions used by Maddison to fill in for missing data, especially during periods of crisis; others stem from inadequacies of the raw source data that can be surmounted by using alternative sources, including recent projects in some countries to develop national accounts for longer periods. Barro and Ursúa also point out that data on consump­tion are an important complement to data on output for various analytical purposes, including the study of asset pricing. Moreover, the dynamics of consumption can vary sharply from the dynamics of output, especially in wartime, when government outlays typically expand while consumption is often restricted.

Altogether, the paper presents nearly full annual data on GDP for thirty-six countries and on aggregate consumption for twenty-four. Defining a “crisis” as a decline of at least 10 percent, the paper identifies 152 crises for GDP and 95 for consumption. The average crisis lasts about 31⁄2 years and represents a decline of more than 20 percent. Crises in GDP and con­sumption generally occur together, with the proportional decline in con­sumption similar to that in output during nonwar crises but larger during wartime crises.

Barro and Ursúa use these data to determine whether the frequency and magnitude of economic disasters justify, for a reasonable assumption of the degree of aversion to risk, the observed premium in returns on equities relative to safer assets. They conclude that the observed average premium of roughly 7 percent on leveraged equity can be justified by a coefficient of relative risk aversion of 3.5, which they deem to be reasonable. This result is robust to various changes in specification with one exception, namely, when the sample is limited to nonwar crises, which eliminates most of the largest declines in consumption.

In the fifth paper, Daron Acemoglu, Simon Johnson, Pablo Querubín, and James Robinson confront the question of why economic policy reforms recommended by analysts sometimes succeed and sometimes fail. A range of reforms urged on developing countries in recent decades— including the opening of markets to trade, financial liberalization, privati­zation of state enterprises, and central bank independence—often have improved economic performance by less than many advocates predicted. One possible explanation, the authors note, is that the reforms were mis­guided, at least when applied to countries at certain stages of development. However, the authors contend that the largest obstacle to improved eco­nomic performance is frequently the political economy of the nation attempting the reform.

Acemoglu and his coauthors argue that determining whether policy reforms will be effective requires understanding why inappropriate poli­cies are being pursued in the first place. They develop a model in which policy distortions arise from a political equilibrium. In this situation, the beneficial effects of distortion-reducing reforms may be offset by the efforts of politically powerful people to recoup losses they suffer when the distortions are reduced. Given the assumptions of this model, policy reform is less likely to be effective in two opposite sets of circumstances: when constraints on political officeholders are so weak that any reform is likely to be undermined, and when the constraints are so strong that policy is not very distortionary initially.

The paper also presents some empirical evidence on one specific type of reform, central bank independence, that the authors view as consistent with these theoretical predictions. Their data suggest that reforms that increase central bank independence reduce inflation more in countries where policymakers face intermediate constraints on their power than in countries where policymakers face strong or weak constraints.

The final paper in this issue is an examination by Niall Ferguson of the behavior of financial markets during the periods surrounding world wars. Ferguson, a historian, begins by observing that despite military interven­tions and terrorist actions and threats, financial markets in this decade have evinced little concern about international political risks. While acknowl­edging that current geopolitical risks appear smaller than those preceding the major world conflicts of the last century, Ferguson contends that threats from the Middle East and Asia and from terrorists around the world have the potential to inflict great harm on Americans and on American property. Moreover, events of the past century showed clearly that interna­tional political crises can have profound effects on financial markets.

Ferguson then reviews the performance of different asset classes around the time of the First World War, the Second World War, and two episodes during the Cold War. He demonstrates that investors did not anticipate the First World War or the effects of the liquidity crisis it caused, but also that asset prices moved very differently in the Second World War, so that investors could not protect their wealth by studying the financial effects of previous crises. As he notes, the chief challenges for investors in learning from the experience of a major war are that such wars are infrequent, and that military and economic paradigms often change dramatically between wars. Ferguson argues that financial markets have become less sensitive to geopolitical risks in recent decades, and he views this evolution as similar to what occurred in the decades between 1880 and 1914. As a result, a major conflict today would hit financial markets “like a bolt from the blue.”