Downloaded From

Germany in an Interconnected World Economy

EDITOR Ashoka Mody

INTERNATIONAL MONETARY FUND

Cataloging-in-Publication Data Joint Bank-Fund Library

Germany in an interconnected world economy / editor, Ashoka Mody. – Washington, D.C. : International Monetary Fund, 2013. p. : ill. ; cm.

Includes bibliographical references.

1. Germany – Economic conditions. 2. Economic development – Germany. 3. Germany – Foreign economic relations. 4. Financial crises – Germany. 5. Labor market – Germany. I.Mody, Ashoka. II. International Monetary Fund.

HC286.G47 2013

Disclaimer: The views in this book are those of the authors and should not be reported as or attributed to the International Monetary Fund, its Executive Board, or the governments of any of its members.

ISBN: 978-1-61635-424-4 (paper) ISBN: 978-1-47558-239-0 (ePub) ISBN: 978-1-47552-249-5 (Mobipocket) ISBN: 978-1-47551-671-5 (Web PDF)

Please send orders to: International Monetary Fund, Publication Services P.O. Box 92780, Washington, D.C. 20090, U.S.A. Tel: (202) 623-7430 Fax: (202) 623-7201 E-mail: [email protected] Internet: www.imfbookstore.org

©International Monetary Fund. Not for Redistribution Contents

Foreword v Christian Kastrop Preface ix Ashoka Mody

1 Tests of German Resilience ...... 1 Fabian Bornhorst and Ashoka Mody The Postwar Catch-Up ...... 3 The Slowdown ...... 5 Reemergence ...... 9 The Great Recession...... 19 References ...... 32 2 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters ...... 35 Martin Schindler Introduction ...... 35 Background: Concepts and Related Literature ...... 38 Methodology and Results ...... 41 Germany’s Growth Sources through a Growth Accounting Lens...... 47 Conclusion ...... 50 References ...... 51 Appendix ...... 52 3 German Productivity Growth: An Industry Perspective ...... 55 Hélène Poirson Introduction ...... 55 German and United States Productivity: Stylized Facts ...... 60 An Industry Perspective ...... 64 Conclusion ...... 72 References ...... 73 Appendix ...... 76 4 What Does the Crisis Tell Us about the German Labor Market? ..... 77 Martin Schindler Introduction ...... 77 Background ...... 78 Recent Developments ...... 80 Understanding German Labor Market Dynamics ...... 85 Conclusion ...... 92 References ...... 94 Appendix ...... 96

iii

©International Monetary Fund. Not for Redistribution iv Contents

5 Growth Spillover Dynamics: From Crisis to Recovery ...... 97 Hélène Poirson and Sebastian Weber Introduction ...... 97 Growth Linkages and Spillovers: Related Literature ...... 100 Empirical Approach ...... 105 Results ...... 109 Channels of Growth Spillover Transmission ...... 123 Conclusion ...... 130 References ...... 131 Appendix ...... 133

6 Do Fiscal Spillovers Matter? ...... 149 Anna Ivanova and Sebastian Weber Introduction ...... 149 Literature ...... 151 Framework...... 155 Simulation Results ...... 158 Conclusion ...... 186 References ...... 187 Appendix ...... 189

7 Current Account Imbalances: Can Structural Policies Make a Difference? ...... 199 Anna Ivanova Introduction ...... 199 Literature Review ...... 202 Baseline Model...... 205 Structural Policies and the Current Account ...... 208 Long-Standing Structural Differences and the Current Account ...... 212 Interaction of Structural Factors and Fundamentals ...... 215 Implications for Germany ...... 217 Conclusion ...... 220 References ...... 221 Appendix ...... 223

8 Discussion ...... 239 Comment on Chapters 2 and 3 ...... 239 Malte Hübner Comment on Chapter 4 ...... 242 Werner Eichhorst Comment on Chapter 5 ...... 247 Felix Hüfner Comment on Chapter 7 ...... 252 Carsten-Patrick Meier References ...... 256

About the Contributors ...... 257

Index ...... 261

©International Monetary Fund. Not for Redistribution Foreword

Germany in its postwar history has a remarkable economic story to tell, not just about success but also about failure, about ground won and lost. The successful post-war recovery, although driven by a clear free-market confession, has also been part of the German policy approach of Ordnungspolitik, which includes a clear commitment to economic fairness, social safety nets, independent wage negotiations between employers and employees, and strong regulations preventing the misuse of economic power in the most economically relevant areas. Then we saw a first slump, starting in the mid-1960s, due to global developments, such as the crises in the coal and steel industries and the loss of the Bretton Woods currency regime at the end of the decade. The slump was prolonged by home-made problems, not least when trying to keep non-competitive sectors alive with high subsidies. Today coal is still active in Germany despite high production costs, and nobody knows if all plants really will be shut down as planned at the end of this decade. Renewables, also heavily subsidized, might not do the trick alone after the closing of all nuclear power plants soon after the Fukushima event. The 1970s brought more trouble with the oil and wage shocks, and they saw the failure of an anti-cyclic macro policy that was also (mis-)used as a tool for goodies of all kind; the cyclical deficits were never paid back in the good times. Last but not least, the decade saw the huge build-up of an over-generous social safety net, at first supported by high growth and stable population figures but soon becoming fragile as birth rates and potential growth rates began shrinking— something that has continued to this day. And the issue hasn’t left us to this day: the question of how to push potential growth is still the most relevant. The 1970s also, as a result, had to manage a level of deficit never seen until the late 1960s, driven by following the new gross investment “(‘golden’) fiscal deficit rule”: that is, the Keynesian-motivated new debt rule, which replaced the old conservative rule that required any deficit to be repaid by revenues created through investment. The new, weakened focus of indirect repayment through growth and taxes, while it may not be completely wrong in economic terms, proved to be a seed of failure. It too could easily be misused, not only as an instrument of political compromise of any incoming new term of legislation but also during economic downturns, when cyclically short-term-communicated deficits “surprisingly” turned out to be structural—the classic TTT (timely, temporary, targeted) mistake mentioned in every economic textbook. The 1980s started with stagflation and debt that had already accumulated quite high and grew further in spite of the changing tide of economic thinking in the direction of neoclassical/monetarist concepts (Phillips-Curve and rational expectations) which—at least in communication—were reflected in the political economy, where non-Keynesian/Ricardian effects were heralded, culminating in statements like, “The economics driven by rational behavior needs no macro.”

©International Monetary Fund. Not for Redistribution vi Foreword

Nevertheless, despite the strong wording, Germany did not really take up the supply-side economy concept and almost completely missed the structural/social reform period of the 1980s that many other European countries entered, such as the Netherlands and Scandinavia. One cushion Germany had at that time was the relatively favorable export side of its economy, even though the deutsche mark slowly but steadily appreciated, taxing away part of gained competitiveness and making structural reforms even more pressing. Before the situation grew too serious, unification brought a big boom. It was construction-driven, and while it did not lead to a general bubble it led to huge overcapacities, which lowered growth in the second half of the decade. The boom wiped the structural reform agenda off the political Top Ten list. Because of its complex and sophisticated system of (federal) checks and balances, therefore politically somewhat “slow,” Germany’s federal government had to spend its limited available political capital on unification, leaving no room for other issues. Seen from outside, it might have been possible to use unification as a catalyst for reform, but the political economy could never deliver. Political capacity was bound by and led to short-term solutions based on already endangered social systems and even higher debt, instead of fundamental reforms and a tax-based financing of unification. The mid-1990s saw Germany taking over Europe’s “red lantern” on economic indicators. The “German Disease” was on the front pages of tabloids and The Economist. The late 1990s brought changes: wage developments were continuously moderated, both specialized SMEs and big industries cleared their books, and profits went up. The supply side won, the demand side lost. The nucleus of the next export boom was born, as German SMEs became the little champions offering emerging markets the technology they needed. Adding to that were Social Security and labor market reforms, tax relief for corporates, and a whole program for renovating Germany, even if, as measured against the bold proposals of economists, the program didn’t look so convincing and was heavily blurred by political compromise at the time of design. But time would show that it was not just the German overperformance but also the underperformance of competing countries (and therefore a no-longer- appreciating currency) that did the trick for the new German economic (export) miracle. For the average German, the crisis of 2008 and the years following—the worst world crisis since the 1930s—was and still is something he would watch with some surprise on television, but it never came closer. And this time Germany was lucky, with a well designed macro-strategy that also delivered on the psycho- logical side and was firmly anchored by the new fiscal rule, the so-called “debt brake.” It was quite a new experience, too, one never seen in the decades before, when the typically pessimistic but now optimistic (over-optimistic?) Germans kick-started lagging internal demand and consumption at the height of the crisis. This book on Germany by Ashoka Mody and current and former IMF and internationally regarded experts is one of the most detailed reviews of recent German economic history. It not only offers an excellent empirical analysis based

©International Monetary Fund. Not for Redistribution Foreword vii on long-term developments but frames that analysis in the context of our actual post-crisis, globalized world. It points clearly at Germany’s successes, but it also points out the remaining weaknesses as well as those upcoming if the reform momentum is lost. Germany has a role to play, not just in Europe but in the interconnected global economy. This should not be forgotten. And that role is not only for its national economy but for the whole region, its trading partners, and its export/ import markets. Germany’s macro- and microeconomic policy does matter, even if measurable spillovers are small. So there is a point in sustaining its reform momentum and not getting lost again on the macro-, structural, supply-and- demand, and all underlying (micro-) factors. Germany indeed should push the structural reform agenda, for example through public investment, reforming the health and service sectors, removing existing market barriers, speeding up research and development, and bridging the existing market gap in founding new-tech companies. It should also adopt family- and birth-friendly policies together with better approaches to migration. Last but not least, Germany should engage in a full renovation of its education system, while keeping its already well working instruments (“dual education”) that prevent youth unemployment. And, of course, there is the issue of debt and deficit. The new German debt brake, which now is also enshrined in the European fiscal compact, might move the countries concerned to an adequate soundness of public finance and foster long-term sustainability and the quality and efficiency of public finance as necessary complements. The reader might use this book to build up his own picture of Germany. The description and analysis are valuable for rethinking different policy approaches, and for reaching the best solutions (or perhaps the second- or third-best, as poli- ticians of all stripes will most naturally and legitimately ensure) to current and upcoming problems, including the next “black swan” problems, in a world that is interconnected and leveraged. It is my hope that the authors of this book keep an open-minded eye on Germany and continue to offer analyses and advice, since in Germany as with other medium and large sized countries a more inward look sometimes tends to become dominant. This is no longer affordable in a globalized, interconnected world. Nobody can do better alone, not with the real and monetary market world and not with the public sector. Maybe this is one of the main crisis lessons. All the more, then, do we have to strive for credible, convincing, and communica- ble—but most of all problem-solving—economic concepts and strategies, not just on a national but on an international scale . Christian Kastrop Deputy Director-General, Economics Department, and Director of Public Finance, Macroeconomics and Research Directorate German Federal Ministry of Finance

©International Monetary Fund. Not for Redistribution This page intentionally left blank

©International Monetary Fund. Not for Redistribution Preface

The papers in this volume were written by IMF staff. But at all stages, this was a collaborative enterprise with the German authorities. Early versions of the papers were discussed at the Bundesbank, where critical comments and alternative viewpoints were offered. That consultative process culminated in a major— and unprecedented—conference at the Federal Ministry of Finance during the 2011 Article IV Consultation with Germany. Christian Kastrop made that event possible. For that initiative and for the many vigorous conversations, even on controversial issues, I am most grateful to the German authorities. At the conference, a number of senior German scholars joined the discussions as commentators and session chairs. Their comments are included as part of this book. Thanks are due in particular to the session chairs, Christoph Schmidt, Klaus Eckhardt, Ansgar Belke, and Beatrice Weder di Mauro, who kindly took time to moderate the exchange following the papers. The conference closed with a lively panel discussion. Juha Kähkönen, then Deputy Director in the IMF’s European Department, moderated that discussion. The distinguished participants included Markus Kerber, then Head of Department of Fiscal and Economic Policy at the Federal Ministry of Finance (now CEO and Director General of Federation of German Industry, BDI), Thomas Mayer, Deutsche Bank Research, and André Sapir, University of Brussels and Bruegel. Finally, I must acknowledge with much gratitude the contributions of Fabian Bornhorst in making this volume possible. He was legitimately a co-editor but his modesty has prevented him from agreeing to share that credit with me. I also recall with great affection my many colleagues who joined this venture and many other such intellectual and operational adventures. Ashoka Mody

©International Monetary Fund. Not for Redistribution This page intentionally left blank

©International Monetary Fund. Not for Redistribution CHAPTER 1 Tests of German Resilience

FABIAN BORNHORST AND ASHOKA MODY

In the nearly 70 years since World War II ended, the German—earlier the West German—economy has enjoyed dynamic phases, interspersed with periods of slow, even anemic, growth. Germany’s continuing ability to maintain competitive manufactured exports has been crucial to its resilience and dynamism, but it has not always been sufficient. This dependence on exports as the economic driver of the German economy helps in understanding its postwar evolution, including its recovery following the post-Lehman collapse. It also points to the challenges that Germany faces: the calls to do more for the European and global economy and the risk that the rise of manufacturing capabilities in emerging markets will eventually wear down Germany’s stronghold. In this introductory essay, we review Germany’s growth record over four phases and sneak a look into the future. • For about a decade and a half after World War II, until the early 1960s, the economy responded spectacularly by making up lost ground. German companies exploited their traditional advantage in capital and durable goods production for sale to a growing world economy and also to meet pent-up domestic demand. • In the next four decades—until about 2003—economic performance gradually became less impressive. The catch-up potential necessarily waned and the oil shocks of the 1970s raised costs and reduced global demand. By the mid-1980s, with rising wages, West Germany was losing its competitive edge. The unification of West and East Germany in 1990 provided a short boost, but then the German economy again went into a swoon. • Starting around 2004, a strong global economy, combined with reforms that helped Germany regain competitiveness, provided a new opportunity to German exporters. With success came a historically large current account surplus, which was criticized for contributing to global imbalances. But the growth dynamic was summarily interrupted in early 2008 by the onset of the “Great Recession,” when the collapse in global trade also swept Germany.

Fabian Bornhorst is an economist in the IMF’s European Department. Ashoka Mody was deputy director in the European Department and then the Research Department of the IMF when this chapter was written. 1

©International Monetary Fund. Not for Redistribution 2 Tests of German Resilience

• The recovery from the collapse has once again demonstrated German resilience. German companies deepened their export links to the growing economies of Asia, and policymakers have supported stabilization and growth through measures to maintain employment and a sizeable fiscal stimulus. New challenges have to be faced. Some are not unique to Germany. With its reliance on exports, a high savings rate, and a current account surplus, Germany shares similarities with Japan, to the extent that one country can be like another. Looking ahead, like their Japanese counterparts, German exporters face height- ened competition, especially from Asia but also from Emerging Europe. And as in Japan, a rapidly aging population will imply a smaller workforce and changes in savings and investment patterns with far-reaching consequences. But Germany’s challenges arise also from its unique role in Europe: in particular, Germany’s expected contribution to the European recovery and a resolution of the euro area crisis remain controversial. As this preview suggests, throughout the postwar era, (West) Germany has benefited greatly from its relationship with the global economy, but there have been lingering questions whether the German contributions to global and European growth have been commensurate. From its early catch-up phase, Germany’s formidable export engine has been its consistent driver of growth. But with almost equal consistency, Germany has run current account surpluses, with imports lagging exports. As a consequence, despite the size of its economy, Germany’s ability to act as global locomotive has been limited. Germany has contributed to global well-being in important ways—notably through its foreign direct investment and, in particular, the production linkages its companies have built throughout Europe. Nevertheless, the calls on Germany to do more remain vigorous, especially as the European crisis has persisted. The argument in this book is that the German economy has evolved along a particular historical trajectory that tends to reinforce its growth patterns. The innovative manufacturing sector has remained a consistent source of growth, but it has been heavily dependent on external demand. The volatility of external demand, in turn, has caused German GDP growth to be relatively volatile by the standards of advanced countries. Such volatility has likely been a factor in keeping domestic demand growth—especially consumption growth—at surprisingly low levels. As a consequence, the domestically oriented segment of Germany’s economy has lagged. This has been so especially in the production and delivery of services. In turn, this has reinforced the drive for foreign markets. Looking ahead, a more domestically driven growth dynamic will be good for Germany and for the global economy. The key challenge for Germany, then, is to generate new domestic sources of growth. One recent episode—reunification—raised domestic demand, but that boom proved unsustainable as the economic problems associated with unification surfaced. Labor market reforms in response to the ensuing period of stagnation were a bold response, and have proven largely successful. But much of the gain in competitiveness translated into further growth in manufactured exports and

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 3 current account surpluses. Therefore, despite the notable service sector advances achieved since the mid-1990s, a further broad-based impetus to services growth is required. Absent such an effort, German growth will remain constrained, and Germany will tend to run current account surpluses while playing only a modest role in spurring growth elsewhere. Just as the German economy is similar to that of Japan—dependent on exports, running current account surpluses, and generating limited international growth spillovers—the contrast with the United States is marked. The United States is characterized by more reliance on domestic consumption and, as a consequence, persistent current account deficits but high international growth spillovers. Despite popular characterization of the U.S. financial sector’s casino capitalism, both the U.S. GDP and its stock market have been less volatile than in either Germany or Japan. In part, this reflects the greater reliance on domestic consumption, which tends to be more stable than exports. But the United States has also been diversified, with innovations in globally leading technologies giving it an edge in productivity growth, especially in the services sector. In the future, Germany will perhaps draw on approaches to fostering innovation practiced in the United States, while also maintaining its greater emphasis on social safety nets, where lessons may be available from the Nordic countries. In this introductory chapter, we step back from the themes of immediate policy focus to provide a broad overview of German economic growth and international connections over the past half century. The intent is to describe the historical trajectory that has brought Germany to its present balance of strengths and weaknesses, and to use that analysis to explore the best way forward. Much of the work underlying this book and its principal policy conclusions was developed for the IMF’s Article IV consultation with Germany in 2011 (IMF 2011). The book serves to present the more extensive background analysis in an integrated form. This chapter is presented in four main parts, reflecting the four relatively dis- tinct phases of German growth. First, the immediate postwar period witnessed a rapid catch-up with the United States and, importantly, reestablished its export prowess, which has remained Germany’s wellspring ever since. Second, growth slowed starting in the early 1960s, and the oil shocks of the 1970s and unification in 1990 proved to be particularly onerous. Third, the reemergence from this setback on the back of extended global prosperity was aided by wage moderation and broad-based domestic reform. And fourth, the collapse triggered by the Great Recession and the subsequent recovery have brought Germany to a new phase with new challenges to tackle.

THE POSTWAR CATCH-UP Emerging from World War II, West Germany set the pace for much of Western Europe as it embarked on a remarkable catch-up. A massive reconstruction effort was mounted and the depleted capital stock was rebuilt. Growth in those immediate postwar years was in the double digits (Figure 1.1).

©International Monetary Fund. Not for Redistribution 4 Tests of German Resilience

Annual growth 5-year average 12 1973 1991 2004 10

−2

−4

−6 1950 1960 1970 1980 1990 2000 2010

Figure 1.1 Germany’s Growth Performance since 1950 (Percent) Source: German Statistical Office. Note: From 1950–1991 data for West Germany. Data prior to 1970 based on different National Accounts methodology and thus not fully comparable.

Aid flowing to Europe through the European Recovery Program (also known as the Marshall Plan) was an important catalyst in addressing infrastructure bottlenecks and reviving trade. Germany also reclaimed its historic advantages as an exporter of capital and durable goods. Between 1948 and 1950, the pent-up demand for consumption and investment drew in substantial imports, but exports started growing rapidly right from the start, with exceptionally high annual growth rates (Table 1.1). By 1950, exports had exceeded imports, and the German trade surplus established then has persisted, with some ups and downs, ever since. By 1960, West Germany’s shares of world exports and imports exceeded those of the German Reich before World War II (Giersch, Paque, and Schmieding, 1992). Eichengreen (2006) describes this as a period of extensive European growth, with Germany in the vanguard. He defines extensive growth as that which deploys relatively well established technologies: “It is the process of raising output by putting more people to work at familiar tasks and raising labor productivity by building more factories along the lines of existing factories” (Eichengreen, 2006, p. 6). Millions of refugees arrived in Germany and, in the immediate postwar

TABLE 1.1 West Germany’s Foreign Trade (Average annual percentage change) 1948–1950 1950–1960 Import volume 26.8 15.0 Export volume 84.4 16.1

Source: Giersch, Paque, and Schmieding, 1992.

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 5 period, internal labor mobility was also high. Nearly full employment conditions were achieved, with the unemployment rate down to below 1 percent in 1960. In reorganizing production and rebuilding the capital stock, efficiency gains were significant and quickly realized. Several factors facilitated this outcome. Not only was plentiful labor available, the workers were also industrially literate—or were readily trainable through the traditional vocational training systems. Such systems met the need of the moment precisely because the challenge at hand was not to build new widgets but to build known widgets in larger quantities with incremental technical improvements in a learning-by-doing process. At the same time, an extensive network of relationship- based banking systems provided the needed patient capital to finance the investments. This confluence led to modest wage demands, which allowed profitable firms the resources and the confidence to invest in their workforce and in growth. Germany was particularly well suited to take advantage of these conditions. Giersch, Paque, and Schmieding (1992, p. 89) write: “Germany’s traditional strength in the production of capital goods paid off handsomely in the 1950s, when these goods were in particularly high demand on the world market. In addition, the change in relative prices testifies to the improving quality and sophistication of these exports goods.” In a similar vein, Eichengreen (2006, pp. 93–94) elaborates: The country already possessed the relevant range of industries, from coal and steel to transport equipment and electrical machinery…. Small and medium-sized firms competed with legions of other small and medium-sized firms, requiring them to price aggressively and reduce costs in order to survive. In turn this rendered German firms highly competitive on international markets. Exports rose from 9 percent of national income in 1950 to 19 percent in 1960. External conditions were also propitious for German recovery. Investment demand was high throughout Europe, aiding German firms specializing in the production of capital goods. The Korean crisis stimulated demand for capital goods worldwide. And just when Germany’s expanding industrial sector began diversifying into the production of consumer goods, private consumption surged across Europe, reflecting rising incomes and in turn helping to sustain the growth of German exports….Investment and exports were the fast-growing components of aggregate demand, and government and private consumption the slow-growing ones. Thus, while all of Europe did well during this period, Germany did particularly well. Other countries that grew about as rapidly included Austria, which had close economic ties to Germany, and Italy. While Germany was leveraging off an established industrial capability, Italian growth was due to the shift of resources from agriculture to industry.

THE SLOWDOWN The precise timing of the shift is difficult to pin down, but the growth benefits of the German miracle leveled off in the 1960s. The indicators are clear. From an average growth of about 9 percent in the early 1950s, GDP growth fell to about 4.5 percent by the mid-1960s (Figure 1.1). German growth, which had set the

©International Monetary Fund. Not for Redistribution 6 Tests of German Resilience

pace for Europe in the immediate postwar years, now fell below the European average growth rate (Figure 1.2). The rapid process of German catch-up with the United States stalled, and even began a modest reversal in the 1980s (Figure 1.3). The slowdown was to be expected. Growth could not persist at the early giddy levels and, as Germany became richer, the convergence possibilities diminished, while the latecomers started their own catch-up process.

8 10th–90th percentile of OECD economies 7 Germany 6 European economies (excluding Germany) 5

−1 1970 1975 1980 1985 1990 1995 2000 2005 2010

Figure 1.2 Growth in a Comparative Perspective, 1970–2010 (Five year average, percent) Sources: IMF, World Economic Outlook; and IMF staff estimates.

Euro area rangea Ireland 110 Spain, Greece, Portugal France Germany Italy 100 90 80 70 60 50 40 30 20 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 Figure 1.3 Per Capita GDP Relative to the U.S., 1960–2010 (Real per capita GDP to U.S. per capita GDP, in percent, PPP constant prices) Sources: Organisation for Economic Co-operation and Development; and IMF, World Economic Outlook. Note: PPP: purchasing power parity. aMaximum excludes Luxembourg.

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 7

The significance of this phase, therefore, lies in two developments, both of which contributed to the slowdown but had longer-term implications lasting into the present. First, the gains possible from extensive growth tapered off and a new phase of intensive growth created new demands on German firms. Eichengreen (2006, p. 6) defines intensive growth as “…growth through innovation. A larger share of the increase in output is accounted for by technical change, and less by the growth of factor inputs.” Thus, it was not merely a matter of narrowing growth opportunities but also a change in the character of those further possibilities and, hence, the resources needed to exploit them. Second, important changes occurred in the organization of the labor market. The rapid growth drew in much of the surplus labor, and with near full employment conditions, the labor market started tightening by the early 1960s. West Germany went from a state of capital shortage to one of labor shortage, which, despite the pursuit of active labor immigration policies (Gastarbeiter), led to demands for higher wages. Reduced profits implied lower investment and growth. Moreover, as wages started rising, employment prospects steadily worsened during the 1970s. The response to this conjuncture was central to shaping the period from the 1970s through the late 1990s. Dew-Becker and Gordon (2012, p. 17) write: … policies adopted to fight unemployment had adverse effects on employment per capita (see Nickell, Nunziata, and Ochel, 2005). To deal with individual hardship caused by higher unemployment, governments increased the generosity and duration of unemployment benefits. To limit the increase in unemployment itself, they attempted to regulate layoffs through employment protection legislation (EPL). To spread the available jobs across the population, they resorted to legislation favoring early retirement and shorter hours of work, so-called “work sharing.” (Alesina, Glaeser, and Sacerdote, 2006) New headwinds gathered momentum in the 1970s with the oil crises and the collapse of the Bretton Woods system, which abruptly changed the external environment for West Germany. Much of the industrial world experienced the angst of declining productivity growth as the pool of innovations appeared to run dry. Germany felt the first taste of competition from newly industrialized economies. German growth slowed to an average of 2.5 percent during the 1970s and 1980s. Domestic consumption remained subdued, and as a consequence import demand was low. Despite the greater competition faced by German exporters, current account surpluses continued (Figure 1.4). German unification in 1990 brought further challenges. The integration of the much less competitive economy of Eastern Germany proved to be economically costly and had a long-lasting influence on economic developments. Unemployment increased substantially, and public finances came under pressure. GDP growth in the 1990s averaged just about 1.5 percent per year.1 Unification did bring about a temporary current account deficit because of a boom in pub- licly led investment, and the trade-to-GDP ratio, already on a mild downward path from the mid-1980s, fell sharply (Figure 1.5).

1 That is, about half of the potential growth rate of the U.S. economy.

©International Monetary Fund. Not for Redistribution 8 Tests of German Resilience

10 10th–90th percentile of OECD economies Germany United States 5 Japan China

−5

−10 1970 1975 1980 1985 1990 1995 2000 2005 2010

Figure 1.4 Current Account Balance in Comparative Perspective, 1970–2010 (Three year average, percent of GDP) Sources: IMF, World Economic Outlook; and IMF staff estimates.

100 Germany 90 Emerging and developing economies 80 Advanced economies 70 60 50 40 30 20 10 0 1960 1970 1980 1990 2000 2010

Figure 1.5 Trade Openness, 1960–2010 (Value of imports and exports, percent of GDP) Sources: IMF, World Economic Outlook; and IMF staff estimates.

By the end of the 1990s, Germany was dubbed “the sick man of Europe.” Growth came to a standstill, and the economy underperformed for much of the early 2000s (Figure 1.6). Despite high unemployment, the institutional responses of the previous decades had raised wages to unsustainably high levels, and for the first time in decades Germany lost a competitive edge (Figure 1.7). Indeed, even as the global economy recovered in the early 2000s, Germany failed to respond, reinforcing the view that Germany faced serious problems. Pessimism about

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 9

Difference between actual and forecast (basis points, right-hand scale) 8 GDP growth 1000 GDP growth forecast (one year ago for the current year) 6 800 600 4 400 2 200 0 0

−2 −200 −400 −4 −600 −6 −800 −8 −1000 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012

Figure 1.6 Growth Expecations in Germany, 1992–2012 (One year consensus forecast vs. actual growth, percent) Sources: Consensus Forecasts; IMF, World Economic Outlook; and IMF staff calculations.

4.0 1991–2000 2001–2010 3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0 Austria France Italy Germany Netherlands Spain

Figure 1.7 Change in Unit Labor Costs, Germany and Selected European Countries, 1991–2000 and 2001–2010 (Average annual percentage change) Source: Organisation for Economic Co-operation and Development. economic prospects became widespread, with growth expectations trailing the actual growth rate for several years into the ensuing global boom.

REEMERGENCE The “reemergence” period, from 2004 to 2008, is a relatively short one, more so when compared to the previous period lasting nearly four decades, which we treated above in summary fashion. Yet this short period is noteworthy because of

©International Monetary Fund. Not for Redistribution 10 Tests of German Resilience

the rapid turnaround experienced and because, along with the post-Great Recession recovery, it defined the prevailing sense of German dynamism. The transformation not only reestablished the German economy’s strengths but also brought calls from the international community for Germany to play a more active role as a “global citizen.” However, as German growth accelerated following years of secular decline, Germany developed large current account surpluses— large even by its own historical standards. Thus, while it came to be regarded with new respect, it was also caught in the storm of “global imbalances.” The essays in this book reflect on both the dynamics of this reemergence phase and on Germany’s role in the international economy. We focus on four themes characterizing this period, reflecting the break and renewed sense of confidence as well as the significant continuities. First, manufacturing productivity growth remained solid, and services productivity growth picked up modestly. Second, wage restraint was widespread. Third, global growth buoyancy was key to maintaining export growth. And, to a large extent, current account surpluses were more a consequence of global developments and less of “distortionary” domestic policies. In Chapter 3 of this book, Hélène Poirson shows that productivity performance varied greatly across sectors. Manufacturing productivity continued to grow respect- ably, not just in absolute terms but also relative to international benchmarks. Productivity growth was relatively low in the newly emerging sectors of communications and information technology and was also relatively low in the services sector. These sectoral distinctions are important, because they highlight the challenge ahead. Despite Germany’s manufacturing prowess, the share of manufacturing value-added in German GDP has steadily declined over the past several decades (Figure 1.8). This is not surprising. With the rise of lower-cost manufacturing in the newly industrializing nations of East Asia, manufacturing has played a smaller role in all advanced economies. Germany is remarkable only to the extent that the manufacturing share of total value-added remains somewhat higher than in Japan. Nevertheless, it is salient that Germany’s high manufacturing productivity growth cannot be a dependable source of greater well-being in the future as manufacturing inevitably continues to cede ground to international competitors. In this respect, Poirson’s analysis has an optimistic note. She finds some evidence of a rise in services productivity starting in the early 2000s. However, her analysis does not establish a clear trend. And since the Great Recession created so much dislocation, any trend may only be discernible in a few more years. Poirson offers advice that is in line with that of other students of productivity growth: greater use of information technology in the delivery of services and more impetus to small and innovative service firms through incubation in higher centers of learning and greater access to venture capital. More controversially, she suggests a role for the government in procuring services with a public-good purpose, especially where that may help establish open standards. The second theme of this period is wage restraint. Wage moderation has become central to characterizing Germany—and hence to the policy measures it must adopt in deference to its global commitments. It is the case that German

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 11

40 Germany Japan United States

15 1970 1975 1980 1985 1990 1995 2000 2005 2010

Figure 1.8 Importance of the Manufacturing Sector, Germany, Japan, and the U.S., 1970–2010 (Value added manufacturing in percent of total value added, three year average) Sources: Organisation for Economic Co-operation and Development; and IMF staff calculations. wages have been relatively steady in the past several years (Figures 1.9 and 1.10). Yet, as described above, the restraint followed years of significant wage increases. Thus, even after the restraint, German wages are among the highest in Europe. What is remarkable about the recent years is the rapid rise in wages elsewhere in Europe. The Irish rise is particularly explosive. But wage increases elsewhere in the European periphery clearly outstripped productivity growth in those economies. There is an open question why German wage increases were modest during this phase. One explanation is the introduction of labor market reforms, in

60 140 Average salary per worker 2010 ('000 EUR) 120 50 Percent increase, 1995–2010 (right-hand scale) 100 40 80 30 60 20 40

10 20

0 0

Italy Spain Ireland Austria France Belgium Greece Germany Portugal Netherlands

Figure 1.9 Wage Levels and Wage Growth, Selected European Countries, 1995–2010 Sources: Organisation for Economic Co-operation and Development; and IMF staff estimates.

©International Monetary Fund. Not for Redistribution 12 Tests of German Resilience

45 1991 1996 2001 2006 2011

Figure 1.10 Wage Share in Germany, 1991–2011(Gross wages as percent of GDP) Sources: World Economic Outlook; IMF staff estimates.

particular the Hartz reforms, introduced around 2003 in association with a broader reform package to regain competitiveness.2 Posen (2007) concludes that the key reform was the reduced duration of unemployment benefits. This increased the labor supply, with the effect of dampening wage growth.3 In Chapter 4 of this book, Martin Schindler argues, as do Burda and Hunt (2011), that the remarkable stability of German unemployment during the Great Recession was due not just to the work-sharing schemes, discussed below, but also to the longer-term effects of the Hartz reforms. While a proper retrospective of the Hartz measures and wage moderation must necessarily be undertaken elsewhere, two cautionary notes on the policy measures are worth considering. First, Dew-Becker and Gordon (2012) point out that starting in the mid-1990s, similar reforms were undertaken throughout much of Europe. This was in response to the rigidities that had become increasingly evident and onerous following the efforts to protect employment in the mid-1970s. The Hartz-like reforms that were undertaken throughout much of Europe had the broad effect of raising employment per capita. The essays in Buti (2009), especially that by Boeri (2009), indicate that more competition was also a feature of Italian labor reforms, with the growth in the incidence of temporary workers. That Germany seems to have harnessed these reforms particularly effectively is, in our view, due to a second consideration. German firms were facing

2 Slow German growth in the early 2000s, while the world economy was beginning to embark on a new dynamic phase, led to a concerted policy effort to renew growth. After years of political deadlock on key reform initiatives, the Agenda 2010 reform program agreed in the early 2000s led to wide- ranging changes in the labor market institutions, a reorganization of the economy, and changes to the pension system. 3 Posen is less convinced of the effectiveness of other elements of the reform package, including the so-called active labor market policies (see also Jacobi and Kluve, 2006).

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 13

increasing competition in international markets and were able to reach an accom- modation on wage demands, much in the manner that occurred during the immediate postwar years. The pressure from international competition and glo- balization forced German enterprises to integrate with (Eastern) Europe and Asia in worldwide supply chains. Such links have been somewhat pejoratively described as the evolution of Germany as a “bazaar economy,” with German companies increasingly engaged in “outsourcing” (e.g., Sinn, 2006). However, this was clearly the way forward, both for Germany and for emerging Europe. Why companies in other countries were less successful in using this opportunity and how the differences in the particulars of reforms and the corporate response played out across Europe remains an important topic for further analysis. With wage moderation and despite the high level of wages, our analysis (see IMF, 2011; and Chapter 7) suggests that Germany’s export engine hummed during these years because of two factors. First, Germany specialized in a range of quality products that were in high demand during this period and for which German exporters faced limited competition. German firms have specialized in a large variety of capital goods, consumer durables, and pharmaceuticals, and they enjoy significant world market shares in these products (Figure 1.11). Germany was able to hold market share, allowing exports to ride the global trade wave (Figure 1.12). Second, because Germany was unable to gain global market share, German export success depended crucially on global prosperity. Before the crisis hit in 2008, global GDP and trade, in particular outside the euro area, expanded at an unusually strong pace (Figures 1.13 and 1.14), and Germany was well posi- tioned to take advantage of that growth.

12 CHN 10 DEU USA 8

6 ITA 4 JPN FRA product varieties NLD

2 KOR ESP MYS IND POL World market share in specialized World IRL PRT 0 GRC 0 100 200 300 400 500

Number of specialized product varieties

Figure 1.11 Product Specialization and Market Share, 2007 (Product varieties and market shares) Sources: IMF staff estimates based on Standard International Trade Classification 4 level trade data for 2007. Note: CHN: China; DEU: Germany; ESP: Spain; FRA: France; GRC: Greece; IND: India; IRL: Ireland; ITA: Italy; JPN: Japan; KOR: Korea; MYS: Malaysia; NLD: Netherlands; POL: Poland; PRT: Portugal; USA: United States.

©International Monetary Fund. Not for Redistribution 14 Tests of German Resilience

500 World trade effect Increased market share 400

300

200

100

−100

India Italy China Korea Spain Japan Poland Greece France Ireland Germany Portugal Malaysia Netherlands United States

Figure 1.12 Decomposition of Export Growth, 2001–2008 (Percent increase) Source: IMF staff estimates. Note: Increase in exports between 2001–08, as percent of exports in 2001, decomposed into the effect of world trade growth and that of increased market share, computed with Standard International Statistical Trade Classification 4 level trade data.

7 EME and DCa US UK Eurozone World 6 5 4 3 2 1 0 −1 −2 1979–88 1989–98 1999–2008 2009–11

Figure 1.13 GDP Growth Rates: World and Major Regions, 1979–2011 (Average annual percentage growth). Sources: IMF, World Economic Outlook; and IMF staff estimates. a Emerging markets and developing countries.

These developments paralleled the emergence of so-called intra-European imbalances, whose sources and implications have been generally misinterpreted. German surpluses vis-à-vis other European economies grew through the 2000s, and Germany is sometimes called on to reverse these surpluses by allowing its wages to rise faster. Aside from the fact that there are no direct ways in which policy can cause wage increases, the analysis focusing on German policies as the reason for intra-European imbalances does not take into account competitiveness gains realized by countries outside of Europe. Thus, Germany continued to

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 15

14 World U.S. Euro area

0 1991–1998 1999–2008 2009–2011

Figure 1.14 Import Growth Rates, World, U.S., and Euro Area (Average annual percentage growth) Sources: Netherlands Bureau for Economic Policy Analysis (CPB); and IMF staff estimates. maintain its traditional exports to Europe, where it was able to fend off competition from Asian sources. But while it continued European imports of intermediate goods, especially from Visegrad countries (Czech Republic, Hungary, Poland, and Slovakia), German imports increasingly tilted toward products produced most cost-effectively by China (Figure 1.15), which became Germany’s second most important import partner in 2011, after the Netherlands, overtaking France. In other words, German exports stayed largely insulated from Asian and lower-wage European competition due to Germany’s specialization, but much of

−2

−4 China Visegrad Rest of the ESP, GRC, Rest of United countries World ITA, IRL, PRT EU-27 States

Figure 1.15 Change in German Import Sources, Selected Countries and Regions, 2000–2009 (Change in Germany’s import shares between 2000–2009) Source: IMF staff estimates. Note: Visegrad countries include Czech Republic, Hungary, Poland, and Slovakia.

©International Monetary Fund. Not for Redistribution 16 Tests of German Resilience

advanced Europe—including the periphery—faced the new reality of global low-wage competition, including when selling to Germany. This suggests that even if higher German wages were to reduce German exports to Europe, those exports would most likely be replaced by imports from Asia, thus making little dent in the European current account deficits. There is the more complex issue of Germany’s current account surplus vis-à-vis the rest of the world. Here there are two separate questions: First, why did the surplus rise to such high levels by 2007–8? Second, even absent that rise, is the surplus too large by some benchmark? In Chapter 7, Ivanova deals with both these questions. On the first, she concludes that the rise in surplus was mainly cyclical. In other words, there was a parallel increase in Chinese, Japanese, and German surpluses that largely mirrored the increased U.S. deficits. The point she makes is that these shifts occurred over a short period of time and so could not principally reflect structural (e.g., labor market policies, regulation) causes. More precisely, her econometric analysis is unable to attribute the rise in the German imbalances to particular policy choices. Ivanova does note that as German exports boomed, corporate profits rose handsomely but investment stayed moribund. Thus, there was a big increase in the domestic savings-investment imbalance, which mirrored the current account surplus. Corporate profits also rose elsewhere, for example in China and Japan. But in China and, to a lesser extent, in Japan, investment also picked up. Why investment in Germany remained so low during these buoyant years is a mystery, although demographic trends combined with the prevailing pessimism about the viability of the German business model in the early 2000s may point to the answer (IMF, 2012). On the second question—whether the German external surplus is too large by some benchmark—Ivanova does find that Germany’s current account surplus implied by fundamentals is somewhat smaller than its observed surplus. But she cautions that the difference is small and the policy measures to engineer it are imprecisely known. That said, she concludes that measures to raise both domestic investment and employment generation (and hence consumption) would be the right way to go. There remains the longer-standing question of Germany’s low consumption growth. After the early postwar years of satisfying pent-up demand, consumption growth tended to fall along with GDP growth. Moreover, as Figure 1.16 shows, there was an inverse relationship between consumption growth and the unemployment rate. In the 1970s, unemployment became a structural problem, and during every downturn until the mid 2000s the unemployment rate rose significantly, whereas subsequent recoveries did not see meaningful reductions. The depressed labor market outlook lowered consumption growth, and by placing an additional burden on the fiscal sector it limited the scope for fiscal policy to stimulate domestic demand (Figure 1.17). With weak domestic demand, the role of exports in generating growth became particularly important in the 2000s. And, while incomes rose, so did household income uncertainty. Despite GDP growth, unemployment remained stubbornly

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 17

12 05 06 10 07 04 08 98 00 99 03 97 95 85 87 8 02 96 01 88 86 09 10 11 9484 89 93 90 6 83 92 91 82 78 4 81 77 80 79 76 Unemployment rate 75 2 74 72 73 71 0 –2 0 2 4 6 Private consumption growth rate

Figure 1.16 Unemployment and Consumption in Germany since 1971 (Unemployment rate and annual private consumption growth rate) Sources: IMF, World Economic Outlook; and IMF staff estimates.

14 Germany 12 West Germany 10

0 1950 1960 1970 1980 1990 2000 2010

Figure 1.17 Unemployment in Germany since 1950 (Unemployment rate, percent) Source: German Statistical Office. high, amplified by the uncertainty over the outcome of labor market reforms. At the same time, the unification shock and specialization of the manufacturing sector toward external markets raised growth volatility. Growth volatility had been lower in Germany than in the United States until the end of the 1980s, but since then it has been typically higher—and even higher than in Japan since the crisis (Figure 1.18 and Carare and Mody, 2010). High GDP volatility in Germany is therefore partly the result of the growth model Germany pursues. Income is volatile, the level of uncertainty is high, saving ratios are higher, and consumption growth is low. These dynamics are

©International Monetary Fund. Not for Redistribution 18 Tests of German Resilience

4.0 Germany United States Japan 3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0 1975–1979 1980–1984 1985–1989 1990–1994a 1995–1999 2000–2004 2005–2008 2009–2011

Figure 1.18 Output Volatility, Germany, U.S., and Japan (Average of 20 quarter rolling standard deviation of year-to-year GDP growth) Sources: Organisation for Economic Co-operation and Development; and IMF staff estimates. a German unification falls in this period.

reinforced by a longer-standing risk aversion, reflected in relatively high stock market volatility (which has consistently been higher than in the United States) and relatively low short-term risk-free rates (which have typically been lower than in the United States except for a brief period after unification and in the most recent crisis period) (Figures 1.19 and 1.20). Similar factors apply to Japan. Weitzman (2008) suggests that persistent high stock market volatility and low interest rates are a reflection of deep consumer uncertainty about structural parameters of the economy.

35 Germany United States Japan 30

0 1993–1997 1998–2002 2003–2007 2008–2011

Figure 1.19 Stock Market Volatility (Five year average of the VDAX, VIX, and VNKY) Sources: Datastream; IMF staff estimates. Note: VDAX: volatility index for the DAX (Germany’s stock index); VIX: volatility index for Standard and Poor’s 500 index; VNKY: volatility index for Nikkei index, Japan.

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 19

Germany United States Japan 14

0 1980–1984 1985–1990 1990–1994 1995–1999 2000–2004 2005–2009 2010–2011

Figure 1.20 Short-Term Interest Rates since 1980, Germany, U.S., and Japan, 1980– 2011 (Yield on Government paper with residual maturity of one year) Sources: Haver Analytics; IMF staff estimates.

THE GREAT RECESSION Although Germany initially took a big hit, the recovery has strengthened its reputation for resilience. With that success have come new demands on Germany to play a more active role in resolving the problems of peripheral Europe. In this section, we describe the German recovery, the policy measures that contributed to it, the interaction of those measures (especially the incentives for work sharing) and corporate strategies, and, finally, the possibilities and limits of Germany’s ability to help beyond its borders. Four of the six subsequent chapters in this book cover themes that deal with the analysis of this period and the policy approaches that are implied.

The Recovery In the immediate wake of the crisis, the fall in German output was substantial, but the recovery was impressive. The decline in output, at over 5 percent, was greater than in the United States and in France and about the same as in the United Kingdom (Figure 1.21). Only with respect to Japan was the fall less severe. Note that both the United Kingdom and Japan have continued to underperform, as if their initial contraction represented, in part, some longer-term downward shift. German exports fell by 14 percent, and German investment also fell pre- cipitously. However, by the end of 2011, Germany had not only recovered past its precrisis output level but, despite its greater initial fall, the German excess over precrisis GDP was greater than in the United States or France. It is important to recognize that the German output recovery impresses mainly because it compares favorably with even weaker performances elsewhere in the advanced world. Almost four years after the start of the crisis, the German output level is now only a few percentage points above its precrisis levels.

©International Monetary Fund. Not for Redistribution 20 Tests of German Resilience

Japan France Germany U.K. United States 104

102

100

88 2008 2009 2010 2011

Figure 1.21 GDP during the Great Recession, Germany and Selected Economies (Real GDP, 2007=100) Sources: IMF, World Economic Outlook; and IMF staff calculations. Germany has recovered, but it would be a mistake to characterize Germany as “booming.” The slow pace of the recovery in Germany and elsewhere and the persistent sense of crisis in the eurozone have meant that this period continues to be described as the Great Recession. With respect to employment, the German story is more remarkable. Employment did not fall much and is significantly above its precrisis levels (Figure 1.22). Forecasters, including the IMF and the German authorities, did not see this coming. In mid-2009, when the unemployment rate stood at about 9 percent, the projections were pointing to a continued rise into double-digit rates. In retrospect, the German performance is noteworthy not only in comparative terms (with other major advanced economies still struggling well below their precrisis employment levels) but also in absolute terms; the gain in employment relative to that before the crisis is a major achievement. Three factors stand out in explaining the German recovery. First, active and coordinated policy measures taken across the globe were crucial in preventing a free fall and fostering a resumption of growth. Eichengreen and O’Rourke (2010) highlight the fact that the initial output collapse in what was subsequently dubbed the Great Recession was, in fact, more severe than in the Great Depression of the 1930s. They attribute the more rapid pace of recovery this time around to a keener perception of the danger of inaction. With the G-20 economies acting in a rare moment of policy coordination to support the global economy, the significantly greater monetary and financial stimulus on this occasion prevented another Great Depression. There was equally a major

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 21

Japan France Germany U.K. United States 104

103

102

101

100

94 2008 2009 2010 2011

Figure 1.22 Employment during the Great Recession, Germany and Selected Economies (2007=100) Sources: IMF, World Economic Outlook; and IMF staff calculations. effort to backstop the financial sector in order to prevent its meltdown. Germany played its role in the coordinated action, through both its financial sector operations and its fiscal stance. In the financial sector, lifelines were provided to banks through the newly constituted SoFFin.4 Although there were some initial concerns that the German authorities might be reluctant to use fiscal stimulus, the German stimulus between 2009 and 2011 was above the average of that in the advanced G-20 economies and only modestly less than that in the United States or Japan (Figure 1.23). Indeed, as Krugman (2010) pointed out, German government consumption between 2007:Q4 and 2010:Q2 rose at a significantly faster clip than in the United States. It may well be that such expansion of consumption had a greater multiplier effect than the tax cuts did in the United States.5

4 The Sonderfonds Finanzmarktstabilisierung (SoFFin) was created in late 2008 to stabilize the financial system by providing bank guarantees, funds for bank recapitalizations, and the transfer of risky assets by creating “bad banks” (IMF 2010). 5 Allen (2005) argues that there has been a historical aversion in Germany to Keynesian demand- oriented remedies in favor of a policy more driven by the goal of expanding the economy’s supply capacity (see also Carlin and Soskice, 2009). Yet, this has also meant a significant social safety net, which provides so-called automatic stabilizers to protect the vulnerable and thereby operates to stabilize the economy. Moreover, as this crisis showed, further discretionary stimulus is pragmatically used in Germany, even if it is often downplayed in public discourse.

©International Monetary Fund. Not for Redistribution 22 Tests of German Resilience

Italy Japan Korea France Canada Australia Germany

United States G-20 Advanced United Kingdom Figure 1.23 Fiscal Stimulus During the Great Recession, Germany and Selected Countries (2009–2011, cumulative, percent of GDP) Source: IMF, Fiscal Monitor, November 2010.

Second, some parts of the world, in particular Asian economies, recovered quickly, and their demand boosted Germany’s exports. China’s investment-driven growth was a major source of world growth (as the advanced world struggled) and played directly to Germany’s strength. Chinese imports of German goods ranged from cars to high-speed rail. And although the Chinese share of German exports was small before the crisis, its incremental contribution to German exports during the recovery phase was substantial (Figure 1.24). China advanced to become

50 Total exports Exports to China 40

−10

−20

−30 2005 2006 2007 2008 2009 2010 2011

Figure 1.24 Growth of German Exports to China, 2005–2011 (Average annual percentage change) Sources: Deutsche Bundesbank; and IMF staff estimates.

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 23

180 Private consumption 170 Employment 160 150 140 130 120 110 100 90 80 1970 1975 1980 1985 1990 19952000 2005 2010

Figure 1.25 Employment and Consumption, Germany, 1970–2011 (Levels, 1970=100 and 1991=100) Sources: IMF, World Economic Outlook; and IMF staff estimates.

Germany’s sixth most important export destination country in 2010, almost on par with the United Kingdom and Italy. Finally, the labor market played a key role. Its resilience is important in itself—both for minimizing the social costs and for the consumption gains it brought to support the economy. However, looking ahead, the strength of the consumption trends remains unclear. The recovery has been driven primarily by exports, especially to areas beyond the eurozone. In turn, exports have pulled up investment. The steadiness of employment notwithstanding, the contribution of consumption to growth has been erratic from quarter to quarter; over the period 2009–2012:Q2 it has been about 30 percent. This is not a surprise. As Figure 1.25 shows, the relationship between employment and consumption growth has weakened since reunification. Note also from Figures 1.18 and 1.20 that GDP and stock market volatility have risen sharply during the Great Recession, with the United States still the lowest, and Germany and Japan maintaining their historically higher uncertainty. Thus, while incomes have grown, so has uncertainty, with the net effect yet to be resolved.

Labor Market Performance The factors behind the strength of Germany’s labor market are discussed in this volume by Schindler (Chapter 4). He concludes that the outcome is a mix of at least four different factors. First, much attention has been paid to government subsidies to maintain employment while reducing the number of hours worked. In Germany, this was implemented through the Kurzarbeit scheme. When the

©International Monetary Fund. Not for Redistribution 24 Tests of German Resilience

crisis hit, the government increased both the duration and the coverage of subsidies. This allowed work to be shared and human capital to be maintained.6 Boeri and Bruecker (2011) point out that similar approaches were adopted in a number of countries, but with less success. They argue that complementary institutions are needed to make this policy effective, pointing especially to employment protection legislation and collective bargaining. Schindler, in Chapter 4, points to a second factor. He notes that the strategy and response of German firms was important. Firms and workers had prior agreements that they could deviate from collectively bargained work arrangements to avoid layoffs, and introduced work-time accounts at the firm level. For workers in the core labor market, this was a reasonable approach: in the face uncertainty in export markets, they opted for job security and flexibility at the firm level. Thus, work-sharing schemes were already a part of the relationship between German firms and their workers. Burda and Hunt (2011) note, moreover, that because workers had credits on these accounts, firing them in the midst of the downturn would have required compensating them for the credits. In this set- ting, the subsidies were helpful in reinforcing the preexisting contractual relationship. Third, as Schindler concludes in Chapter 2, for Germany the Great Recession was mainly an external export shock, not a supply shock. Despite the uncertainty, firms largely saw this as a temporary shock and relied on accumulated work time and short work schemes instead of layoffs. With the recovery, this judgment was vindicated. Burda and Hunt (2011) argue that the precrisis expansion of exports was viewed by firms as temporary, so firms held back their hiring—this is consistent with the relatively low level of corporate investment in the boom years, as Ivanova has noted. Thus, firms did not have as much of a need to fire people. Looking beyond the near term, Schindler in Chapter 4 suggests that German unemployment had begun a secular downward trend prior to the crisis, possibly triggered by the Hartz reforms. The evidence for this conclusion is still preliminary, although it is plausible since these reforms increased labor flexibility both among the core work force and also at its margins. Once again, it helps to look beyond Germany. As noted above, other European countries adopted similar labor market reforms in the mid-1990s. Boeri (2009) finds that similar reforms to the Italian labor market led to favorable labor market responses in that country: he estimates that the level of unemployment consistent with non-accelerating inflation came down. However, Boeri goes on to argue that Italian firms’ essential lack of dynamism implied that employment did not respond in meaningful numbers. Here again, Germany’s historical strengths and corporate-labor relationships distinguish it from other European nations.

6 This would be of particular importance if skilled labor was in short supply. In chapter 3 of this volume, Schindler notes that the firms’ more intensive use of the core labor force, combined with an adjustment in the temporary workforce (typically less skilled) is consistent with such considerations.

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 25

Germany’s International Role Germany’s economic size and its recovery from the Great Recession have led to calls for more German support to global growth, especially to the euro area. We do not deal here with the broader issues of euro area governance and financial arrangements to support “bailouts.” Rather, our focus is on the specific role that German economic policy can play in fostering growth and stability in the eurozone. In this context, three sets of issues arise7: • Germany as a growth “locomotive,” • German fiscal policy as stimulus for Europe, and • The support provided to peripheral nations through the Target 2 system. Germany’s ability to act as either a global or even a European locomotive is limited. This follows from the theme that this essay has developed, namely, that German growth has itself depended to a large extent on global growth. While there is some optimism that growth will become more domestically driven with structural improvements in the labor market, the evidence for that, as discussed above, is at best preliminary. For a nation to act as a growth locomotive for others, it must originate growth impulses. This, in essence, is the finding of Hélène Poirson and Sebastian Weber in Chapter 5 of this volume. Growth spillovers from Germany to the rest of the world remain limited because domestically originated growth is limited. Rather, Germany acts as a transmitter of global trade impulses, mainly from the United States and Asia to Europe. German supply chains that seek inputs for the delivery of Germany’s own exports extend to its eastern trading partners. But these supply chains are set in motion by demand that predominantly originates outside of Europe. For Germany to assume more of a locomotive role would require autonomous sources of demand from within the country. There are no easy policy steps to achieve this outcome. Ultimately, efforts to raise German growth potential in a manner that raises labor participation and emphasizes the production and delivery of services are likely to raise both German economic welfare and the contribution to global growth. But what if Germany used its fiscal space to stimulate demand for goods and services produced in the periphery? Would that not provide much needed temporary relief? The answer is, in principle, yes. However, Anna Ivanova and Sebastian Weber in Chapter 6 of this volume conclude that the quantitative effect would be small. Spillovers from an activist fiscal policy in Germany to the periphery are small, because trade links are weak. The chapter shows that the arithmetic works against fiscal spillovers, because the quantitative effect is further reduced if the German domestic multiplier is less than one, and is further diluted since only a

7 We have addressed above the issue of intra-European imbalances. As noted, a policy that weakens German competitiveness may reduce German surpluses but not necessarily improve the current account positions of the peripheral countries.

©International Monetary Fund. Not for Redistribution 26 Tests of German Resilience

800 Germany Netherlands

600 Belgium France Italy Greece, Ireland, and Portugal 400 Finland Luxembourg 200

−200

−400 Jan 05 Jan 06 Jan 07 Jan 08 Jan 09 Jan 10 Jan 11 Jan 12

Figure 1.26 Target 2 Balances, Germany and Selected Euro Area Countries, 2005– 2012 (EUR billion) Sources: IFS; National Central Banks; and staff estimates.

share of domestic spending is diverted to imports, of which only a small share benefits the periphery. While support to European growth, either through general economic activity or through fiscal policy, is likely to be limited, as part of the Eurosystem Germany has somewhat unwittingly played an important role in providing financial stability and hence preventing a more serious growth collapse. This has occurred through the so-called Target 2 system. Target balances are the mechanism through which national central banks within the Eurosystem lend to each other. Target positions are intended to smooth over temporary liquidity needs of the member countries, and the system has typically been close to balance. Since 2007, however, imbalances have grown steadily, and as of this writing Germany has a large creditor position (Figure 1.26), while a number of countries have become net debtors (Bornhorst and Mody, 2012). In effect, as the crisis deepened, capital flows in the eurozone reversed. The creditor countries (especially Germany and the Netherlands) opted not to roll over their financial exposure to much of the rest of the eurozone (especially to the most stressed economies), including to the cross-border interbank market. This was yet more evidence that the flows in the first place had not generated productive investments. The process gained momentum in the fall of 2011, and with the debtor nations unable to meet these repatriation obligations, the only vent to ease what would have been unbearable pressure was the ECB’s Target 2 system (Figures 1.27 and 1.28). Essentially, the ECB mediated excess balances in the creditor central banks to banking systems of the stressed economies.

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 27

600 Financial flows ex target (+ outflow) Change in target claims (+ increase) 400 Current account

200

−200

−400 2005:Q1 2006:Q1 2007:Q1 2008:Q1 2009:Q1 2010:Q1 2011:Q1 2012:Q1

Figure 1.27 Current and Financial Account of Germany, Finland, and the Netherlands since 2005 (Billions of euros, rolling four quarter sum) Sources: Haver Analytics; IFS; IMF staff calculations.

600 Financial flows ex target (+ outflow) 400 Change in target claims (+ increase) Current account 200

−200

−400

−600 2005:Q1 2006:Q1 2007:Q1 2008:Q1 2009:Q1 2010:Q1 2011:Q1 2012:Q1

Figure 1.28 Current and Financial Account of Greece, Ireland, Italy, Portugal and Spain since 2005 (Billions of euros, rolling 4 quarter sum) Sources: Haver Analytics; IFS; IMF staff calculations.

The Emerging Challenges We end this review with a cautionary gaze into the future. We ask two questions. Can Germany continue to assume leadership in high-end manufacturing as the basis for its economic strength? And, how important could the consequences of population ageing be? Over the past few decades, advanced economies have gradually ceded ground in manufacturing to newly industrializing economies. The initial challenge came

©International Monetary Fund. Not for Redistribution 28 Tests of German Resilience

in labor-intensive products, such as garments, shoes, and consumer electronics, where cheap labor played an important role. But over time, as their wages have risen and their industrial competency has grown, the new competitors have expanded their line of products and raised their quality standards. While the first televisions from Korea came in boxes with Japanese brand names, today Korean televisions can carry their own premium. Similar progress has been achieved in semiconductors and automobiles. The entry of China as the export machine has rendered this process even more dynamic. The gap between emerging and advanced nations is narrowing fast, and know-how is increasingly mobile. Germany has so far warded off this challenge. First, German specialization has been particularly resistant to international competitive challenge. Thus, German producers have continuously built on their strengths to maintain their lead in their product niches. Also, they have recognized the opportunities from the opening up of lower wage economies and brought them in as partners who supply various inputs and assembly services. But will this be sufficient? The United States also enjoys specialization in a large number of products, but it has been losing market share in the past decade. Similarly, Japan has been ceding ground in an increasing range of products. It would be a surprise if even the areas of traditional German dominance do not face stepped-up competition in the next generation. The threat from ageing is well recognized. As this essay has highlighted, Germany and Japan share a number of common features. Japan, however, has aged much faster. As Figure 1.29 shows, the ratio of Japanese over age 65 shot up in the two decades from 1990 to 2010 and now stands at about 22 percent. This has resulted in important changes in the Japanese economy.

24 Germany Japan United States 22

10 1990 1995 2000 2005 2010

Figure 1.29 Population Ageing, Germany, Japan, and the U.S., 1990–2010 (Popula- tion over 65 years, in percent of total population) Source: Organisation for Economic Co-operation and Development.

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 29

Recent research (Braun, Ikeda, and Joines, 2007) shows that the single most important cause of the decline in Japanese savings rates is lower savings as people get older. Yet this has not necessarily meant a consumption boom. Rather, even with the fall in savings, Japan has experienced subdued demand and deflation. The relationship between ageing and deflation is, of course, more tenuous, but it is presumably related to the shrinking workforce (Figures 1.30 and 1.31). Germany is about to go through a more intensive ageing phase, following Japan. The

40 Percent 65+ 2010 Percent 65+ 2050

0 United States Germany Japan OECD

Figure 1.30 Population Ageing, Germany, U.S., Japan and the OECD, 2010–2050 (Population over 65, in percent of total population) Source: Organisation for Economic Co-operation and Development Labour Force and Demographic Database, 2010.

250 Forecast France

Japan Germany 200 UK US

150

100

50 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050

Figure 1.31 Working-Age Population, Germany and Selected Countries, 1950 to Present and Projected (1950=100) Source: United Nations.

©International Monetary Fund. Not for Redistribution 30 Tests of German Resilience

10th–90th percentile of OECD economies Germany

80 Nordic economies (DNK, FIN, SWE, NOR) Average 75 70 65 60 55 50 45 40 35 30 1970 1975 1980 1985 1990 1995 2000 2005 2010

Figure 1.32 Female Labor Force Participation, Germany and Nordic and OECD Countries, 1970–2010 (18–64 years old, percent) Sources: Organisation for Economic Co-operation and Development; and IMF staff estimates.

10th–90th percentile of OECD economies GermanyGermany

Nordic economies (DNK, FIN, SWE, NOR) AverageAverage 9090

8585

8080

7575

7070 1970197019751975 1980 1985 1990 1995 2000 2005 2010

Figure 1.33 Male Labor Force Participation, Germany and Nordic and OECD Countries, 1970–2010 (18–64 years old, percent) Sources: Organisation for Economic Co-operation and Development; and IMF staff estimates.

pressures are rising as the population ages, and migration, even by the most optimistic scenarios, cannot fill the emerging gap in the working-age population. This will call for policies, possibly following the Nordic model, directed toward increasing the labor force participation especially of women (Figure 1.32 and 1.33). This in turn implies adjustments to effective tax rates to enhance female labor supply along with supportive child care policy.

Conclusion: Germany in an Interconnected World Despite economic ups and downs, Germany has held a commanding position in the global economy over the past half century. This position primarily reflects

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 31 an industrial strength that has been tested in global competition over decades and has proven resilient to severe setbacks. Most recently, the period after German unification saw the economy in doldrums with seemingly intractable high unemployment rates. Yet the German economy has continually displayed a capacity to regenerate itself. Its emergence from the Great Recession—and especially the ability to generate jobs—has been widely, and rightly, lauded. In turn, this resilience is due to a network of innovative firms that have adapted to global changes, for example through increased sourcing from Eastern Europe and exploiting new market opportunities in Asia. It is also due to the ability of firms and labor to find common ground. Finally, it is due to a pragmatic bent in policymaking. Following World War II, Germany experienced a particularly robust reconstruction phase. This was followed by an extended period of slowing growth. First, there was a natural slowing after the rapid postwar catch-up. Then the external shocks of rising oil prices and the collapse of the Bretton Woods system caused further deceleration. The unification episode was unique to Germany. After a short-lived unification boom in the early 1990s, growth fell again. When in the 1990s the world economy began a new expansion phase, the German economy appeared ill-prepared to take advantage of the favorable environment. But by the mid-2000s, economic reform and corporate and labor responses to the changed circumstances led to a German reemergence. Since then, the German economy has displayed considerable strength and maturity, not least by robustly navigating the crisis of 2008–9. In particular, German employment levels weathered the recent crisis surprisingly well. With its success, Germany has been called on to assist the global recovery, questions about its current account surplus have resurfaced, and Germany’s wage moderation and proposed pace of fiscal consolidation following the recovery from the 2008–9 crisis have, at times, been regarded with concern. However, while proposals for more rapidly raising German wages or delaying fiscal consolidation have a plausibility, a closer examination suggests that they could compromise German strengths with dubious short-term stimulative value for other countries. The real German challenge is to strengthen its areas of weakness. Although significantly down from their peak levels, unemployment rates remain elevated, including when judged by Germany’s own past history. These high rates, along with the emergence of relatively low-paying and temporary jobs, also act as a drag on German consumption growth. The key is to counteract medium-term growth constraints in a way that also supports sustainable rebalancing via higher domestic demand growth. Meeting this challenge will require a new generation of pragma- tism in policy decisions. A greater emphasis is needed on innovation that extends Germany beyond its traditional manufacturing strengths and on a new model of social safety nets, drawing possibly on the Nordic experience, to increase labor force participation needed to counter rapid ageing. Such actions would also be good for Europe and the global economy.

©International Monetary Fund. Not for Redistribution 32 Tests of German Resilience

REFERENCES Alesina, Alberto F., Edward L. Glaeser, and Bruce Sacerdote, 2006, “Work and Leisure in the U.S. and Europe: Why So Different?” in NBER Macroeconomics Annual 2005, ed. by Mark Gertler and Kenneth Rogoff (Cambridge, MA: MIT Press), pp. 1–100. Allen, Christopher S., 2005, “Ordo-Liberalism Trumps Keynesianism: Economic Policy in the Federal Republic of Germany,” In Monetary Union in Crisis: The European Union as a Neo-Liberal Construction, ed. by Bernard Moss London (Palgrave), pp. 199–221. Boeri, Tito, 2009, “Comments to ‘Two Italian Puzzles: Are Productivity Growth and Competitiveness Really So depressed?’ by Lorenzo Codogno,” in Italy in EMU: The Challenges of Adjustment and Growth, ed. by Marco Buti (London: Palgrave Macmillan). ———, and Herbert Bruecker, 2011, “Short-time Work Benefits Revisited: Ssome Lessons from the Great Recession, Economic Policy, vol. 26, pp. 697–765. Bornhorst, Fabian, and Ashoka Mody, 2012, “Target Imbalances: Financing the Capital- Account Reversal in Europe,” article published online by VoxEU.org. http://voxeu.org/ article/target-imbalances-financing-capital-account-reversal-europe. Braun, R. Anton, Daisuke Ikeda, and Douglas H. Joines, 2007, “The Saving Rate in Japan: Why It Has Fallen and Why It Will Remain Low,” CIRJE Working Paper F-535, CIRJE, Faculty of Economics, University of Tokyo. Burda, Michael C., and Jennifer Hunt, 2011, “What Explains the German Labor Market Miracle in the Great Recession?” SFB 649 Discussion Paper No. 2011-031, Humboldt Universität Berlin. Buti, Marco, ed., 2009, Italy in EMU: The Challenges of Adjustment and Growth (London: Palgrave Macmillan). Carare, Alina, and Ashoka Mody, 2010, “Spillovers of Domestic Shocks: Will They Counteract the ‘Great Moderation?’” IMF Working Paper No. 10/78 (Washington, DC: International Monetary Fund). Carlin, Wendy, and David Soskice, 2009, “German Economic Performance: Disentangling the Role of Supply-Side Reforms, Macroeconomic Policy and Coordinated Economy Institutions,” Socio-Economic Review, Vol. 7, pp. 67–99. Dew-Becker, Ian, and Robert J. Gordon, 2012, “The role of Labor Market Changes in the Slowdown of European Productivity Growth,” Review of Economics and Institutions, Vol. 3, No. 2, pp. 1–45. Eichengreen, Barry, 2006, The European Economy Since 1945: Coordinated Capitalism and Beyond (Princeton: Princeton University Press). ———, and Kevin O’Rourke, 2010, “A Tale of Two Depressions,” article published online by VoxEU.org. http://voxeu.org/index.php?q=node/3421. Giersch, Herbert, Karl-Heinz Paque, and Holger Schmieding, 1992, The Fading Miracle: Four Decades of Market Economy in Germany (Cambridge, UK: Cambridge University Press). International Monetary Fund, 2010, “Germany: Staff report of the 2010 Article IV Consultation” (Washington, DC: International Monetary Fund). http://www.imf.org/ external/pubs/ft/scr/2010/cr1085.pdf. ———, 2011, “Germany: Staff report of the 2011 Article IV Consultation” (Washington, DC: International Monetary Fund). http://www.imf.org/external/pubs/ft/scr/2011/cr11168.pdf. ———, 2012, “Germany: Staff report of the 2012 Article IV Consultation” (Washington, DC: International Monetary Fund). http://www.imf.org/external/pubs/ft/scr/2012/cr12161.pdf. Jacobi, Lena, and Jochen Kluve, 2006, “Before and After the Hartz Reforms: The Performance of Active Labour Market Policy in Germany,” IZA Discussion Paper No. 2100. Krugman, Paul, 2010, “Ever Expanding Government,” September 10 posting in “The Conscience of a Liberal” blog, New York Times. http://krugman.blogs.nytimes.com/ 2010/09/10/ever-expanding-government/

©International Monetary Fund. Not for Redistribution Bornhorst and Mody 33

Nickell, Stephen, Luca Nunziata and Wolfgang Ochel, 2005, “Unemployment in the OECD Since the 1960s: What Do We Know? Economic Journal Vol. 115, No. 500, pp. 1–27. Posen, Adam, 2007, “Hartz IV Worked—As Far As It Went,” Op-ed in Die Welt, March 14. http://www.iie.com/publications/opeds/oped.cfm?ResearchID=717. Sinn, Hans Werner, 2006, “The Pathological Export Boom and the Bazaar Effect – How to Solve the German Puzzle,” CESifo Working Paper No. 1708 (Munich: CESifo Group). Weitzman, Martin L., 2008, “Subjective Expectations and Asset Return Puzzles,” American Economic Review, Vol. 97, No. 4, pp. 1102–30.

©International Monetary Fund. Not for Redistribution This page intentionally left blank

©International Monetary Fund. Not for Redistribution CHAPTER 2 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

MARTIN SCHINDLER

While the German economy, with a GDP decline of 4.75 percent in 2009, was among the hardest hit during the financial crisis, the impact of the crisis on its potential output is estimated in this chapter to have been comparatively mild and transitory, with virtually no potential output loss in the long term. This outcome reflects both a more flexible economic structure, resulting from past reforms, and, importantly, the nature of the shock—a temporary drop in external demand. The temporary nature of the shock meant that no large-scale structural shifts in employment were necessary, so employers were inclined to retain workers and adjust hours per worker instead. This incentive was buttressed by policy measures during the crisis, but more importantly also by longer-standing labor market improvements regarding hourly flexibility. With employment levels and, especially, hours, returning to pre-crisis levels, the impact on potential GDP was limited, and the evidence suggests that a medium-term convergence of potential output to its pre-crisis trend is likely. However, there is little indication that long-term potential growth will rise above its meager historical rate of about 1.25 percent. Increasing potential growth in the medium and long term will require sustained efforts to reverse the declining secular trend in Germany’s productivity growth and to raise its low total factor productivity (TFP) growth.

INTRODUCTION Germany’s economic structure has changed dramatically over the past four decades. Until about the mid-1990s, the contribution of external trade to aggregate growth—in either direction—was similar to that in most other advanced

The author has benefitted from comments and suggestions by Ashoka Mody, Hélène Poirson, Anna Ivanova and participants at the Germany in an Interconnected World Conference at the Ministry of Finance in Berlin (May 2011), especially the paper’s discussant, Malte Hübner. Roberto Garcia-Saltos, Petar Manchev, and colleagues at the IMF’s Research Department Economic Modeling Unit provided helpful advice on implementing the Matlab code. Susan Becker provided excellent research assistance. 35

©International Monetary Fund. Not for Redistribution 36 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

1.5 1971–1995 1996–2010

1.2

0.9

0.6

0.3

0.0 Germany Japan United States United Euro area Kingdom (excluding Germany)

Figure 2.1 Foreign Contributions to Growth (Annual percentage points) Sources: IMF; World Economic Outlook; and IMF staff calculations.

economies.1 Since the mid-1990s, however, the importance of foreign trade in cyclical fluctuations (in an accounting sense) has increased dramatically in Germany, exceeding that in other advanced economies by far, even as its real GDP growth rate ranked among the lower rates during that period. That is, during 1996–2010, annual growth on average was almost exclusively driven by increases in net exports, with domestic growth contributing a negligible amount (Figures 2.1 and 2.2). The increased importance of external trade also featured prominently during the recent crisis. More so than elsewhere, Germany’s crisis contraction and its recovery were correlated with fluctuations in external demand. (Figure 2.3) In particular, the downturn was markedly different from that in most other economies in that it almost exclusively reflected a decline in net exports: between 2007:Q4 and 2009:Q1, real GDP declined by a cumulative 5.8 percent, of which 5.1 percentage points were due to the cumulative decline in real net exports.2, 3

1 Time-series data exhibit a V-shaped pattern, with the German foreign contribution to growth more elevated in the 1970s and 80s, though still substantially below the more recent uptick in the trade surplus. The increase in the foreign contribution to growth is even more pronounced in relative terms, i.e., the foreign growth contribution relative to total growth. For a more detailed analysis of Germany’s historical growth drivers, see Vitek (2010), which also illustrates the increasing importance over time of foreign demand in German business cycle fluctuations, especially during the Great Recession. 2 Exports of goods and services dropped by over 15 percent in real terms during that period, with the impact on net exports somewhat offset by an import decline of just under 6 percent. 3 The theme of strong external growth drivers alongside relative moderate contributions from domestic sources raises deeper questions regarding their causal connection: has weak domestic demand led firms to focus on external markets? Or has a dominant external sector, e.g., through its increased volatility, led individuals to raise (precautionary) savings and lower their consumption? Or are there unrelated causes for the two? This is an important theme that, however, cannot be addressed in the context of this chapter.

©International Monetary Fund. Not for Redistribution Schindler 37

4.0 1971–1995 1996–2010 3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0 United United Euro area Germany Japan States Kingdom (excluding Germany)

Figure 2.2 Average Real GDP Growth (Annual percentage points) Sources: IMF, World Economic Outlook; and IMF staff calculations.

8 6 Domestic Net exports 4 2 0 –2 –4 –6 –8 –10 France Germany Japan Sweden United United Kingdom States

Figure 2.3 Growth Composition during Crisis and Recovery Sources: IMF, World Economic Outlook; and IMF staff estimates. Note: The bars represent cumulative contributions to growth from domestic and external sources during 2008:Q2–2009:Q1 (left bar) and 2009:Q1–2010:Q3 (right bar).

The reliance on external demand affects how growth shocks are transmitted to potential GDP. Generally speaking, any given change in actual GDP may reflect either a fluctuation of actual GDP around a stable potential GDP path (that is, fluctuations in demand) or a change in potential GDP (that is, fluctuations in supply), or both. The assessment of what type of shock is the source of the observed fluctuations in GDP therefore has strong implications for the assessment of the extent to which potential GDP may be affected.

©International Monetary Fund. Not for Redistribution 38 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

Building on the notion that external demand has been the main driver of German business cycles—at least since the mid-1990s—the analysis in this paper concludes that potential growth was only minimally affected by the crisis, in contrast to many other countries during the crisis (including the United States) and in contrast to the effects typically associated with past financial crises.4 Thus, the demand shock has pulled down actual GDP, but it has not set in motion firm activities that would substantially alter Germany’s growth potential in the medium term, as reflected particularly in the moderate crisis impact on capital stock and employment. Consistent with this, a robust recovery in actual GDP has pulled up potential growth, which has been seen to overshoot its long-term rate in the medium term, making up for the small potential growth losses during the crisis. The key result is that there is likely to be no permanent loss of potential GDP relative to the pre-crisis trend. On a more somber note, however, the underlying assumption of a low long- term potential growth rate of only about 1¼ percent leaves Germany lagging behind many of its peers, especially the United States, which, despite a more severe crisis impact, is set to continue to grow at a faster rate than Germany. Given Germany’s adverse demographic pressures, only continued structural improvements to raise productivity and the supply of labor can lift German GDP to a higher trajectory. This chapter provides some background on the existing literature on potential GDP, including its definition and selected findings; describes the methodology and the main results; and presents these results within the context of a growth accounting exercise that decomposes German growth into that deriving from total factor productivity dynamics, (physical) capital accumulation, and employment growth. An appendix provides further details on the underlying model that was used to estimate potential GDP.

BACKGROUND: CONCEPTS AND RELATED LITERATURE Potential GDP has been defined in different ways in the related literature. Some authors provide an essentially statistical definition, such as Gordon (2008, p. 1) who “uses the adjectives ‘potential’ and ‘trend’ as synonyms to describe the long-run growth rate of real GDP… after cyclical elements of the observed data are factored out by a statistical de-trending procedure.” Others define potential output more conceptually as the full-employment level of output, that is, “the level of output at which the economy’s resources are fully employed” (Mankiw, 2002, p. 246), where full employment is understood to mean that capacity utilization and unemployment rates are at their “normal”

4 For example, Furceri and Mourougane (2009) estimate that financial crises have on average perma- nently reduced potential GDP by 1.5–2.4 percent in OECD economies during 1960–2007.

©International Monetary Fund. Not for Redistribution Schindler 39 or “natural” levels. In the latter definition, potential growth is typically viewed as the maximum growth rate that is sustainable without an acceleration of inflation. The different definitions of potential output are reflected in different approaches to measuring or estimating it. The more statistical definition has the advantage of being easily implemented through various statistical de-trending methods, such as simple averaging or the Hodrick-Prescott filter. Definitions relying on economic concepts require the use of economic models that help to estimate differences between actual and potential output, often based on estimates of equilibrium levels of labor and capacity utilization rates. The approach in this paper falls on the side of the latter approach. Applied to Germany, a simple averaging of annual growth rates suggests a long-term potential growth rate of 2 percent annually during 1970–2011 (Figure 2.4). However, the German reunification represents an important structural break—prior to that, annual growth averaged 2.7 percent, while it has been averaging a lower rate of 1.3 percent during the post-unification period. The secular decline in potential growth is consistent with, but sharper than, Gordon’s (2008) estimates for the United States. Based on simple growth averaging between “benchmark dates” (Figure 2.4), Gordon finds a decline in the United States from nearly 3.5 percent in the early 1970s to about 2 percent in 2007–08. Notably, this decline is much more moderate than in Germany—according to these simple estimates, Germany’s potential growth was broadly on par with the United States in the 1970s and 1980s, but is now substantially lower. That said, the analysis that follows in this chapter suggests that this gap may once again

8 United States averages (Gordon, 2008) Annual 6 Pre-/post-reunification averages

−2 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

−4

−6

Figure 2.4 Germany and the United States: GDP Growth (Percent) Sources: IMF, World Economic Outlook; Gordon (2008); and IMF staff estimates.

©International Monetary Fund. Not for Redistribution 40 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

shrink as the damage that the financial crisis has caused to medium-term potential growth is likely to be more substantial in the United States than in Germany.5 Over longer horizons, such trend-based estimates should conform closely with what would be obtained based on the economic definition of potential output, since on average an economy should not operate at a level substantially different from its potential over long periods. Thus, over long horizons, simple trend estimates are likely to be reasonable estimates of potential output and can serve as inputs in the type of modeling approach used here to anchor the estimates. However, as shown below, more substantial differences often arise over shorter frequencies, where atheoretical methods may be unable to sufficiently differentiate between changes in actual output and changes in potential output. A number of studies have examined potential output in Germany. Based on a production function methodology, De Masi (1997) estimates potential growth in Germany at 2.25–2.5 percent, which is consistent with the long-term growth rate reported above, given her 1980–1997 sample. Also following a production- function approach and considering the 1986–2003 time period, Baghli, Cahn and Villetelle (2006) find a secular decline in potential growth rates in Germany, although their post-unification average of about 1.9 percent is somewhat higher than the average calculated above, reflecting the absence of post-2003 data, especially the low growth rates until 2005 and the more recent crisis (see also Cahn and Saint-Guilhem, 2007). That potential output growth estimates are subject to considerable uncertainty is exemplified by Horn, Logeay, and Tober (2007) who find, for various versions of their model, annual potential growth rates during 2006–2010 of 2.4, 2.1 and 1.3 percent, respectively, owing to different assumptions on total factor productivity (TFP) dynamics and the level of the NAIRU.6 At the lower end are both the Sachverständigenrat (2011) and Deutsche Bank Research (2011) studies, which find potential growth to be on the order of 1.25 percent, at least starting in 2000 (and about 1.5 percent during the 1990s).7

5 The notion that Germany’s potential output held up well during the crisis but lags from a longer- term perspective also holds in a broader cross-country comparison. For example, as the discussion in Section D of this chapter indicates, German potential GDP growth from 2000 until the onset of the crisis fell behind that of the United States, the United Kingdom, Japan, and the Euro Area as a whole. As also shown there, TFP growth, the main driver of a country’s long-term growth potential, is low in Germany compared to other OECD economies. 6 Also, while Horn, Logeay and Tober (2007) utilize data dating back to 1970, they do not explicitly model a structural break at the time of reunification. (NAIRU = Non-Accelerating-Inflation Rate of Unemployment). 7 Consistent with the main message of this chapter, both Sachverständigenrat (2011) and Deutsche Bank Research (2011) expect potential growth to be little affected by the crisis. El-Shagi (2011) comes to a similar conclusion.

©International Monetary Fund. Not for Redistribution Schindler 41

TABLE 2.1 Model Input Data Variable Description Source Real GDP Gross Domestic Product (SA/WDA, Bil.Chained.2000. Statistisches Euros) Bundesamt Core Inflation HICP: Total excluding Energy and Unprocessed Food Haver (SA, 2005=100) Capacity Utilization Harmonized Capacity Utilization: Manufacturing European (SA, percent) Commission Inflation Expectations CPI percent chg. 6–10 years ahead, interpolated to Consensus Forecasts quarterly frequency Unemployment Unemployment Rate (SA, percent) Statistisches Bundesamt

Sources: See third column. Note: SA: seasonally adjusted; WDA: work day adjusted; Bil.Chained.2000: billions of 2000 euros, chained; HICP: Harmonized Index of Consumer Prices; CPI: Consumer Price Index.

METHODOLOGY AND RESULTS Following Benes and others (2010), potential output is estimated here using a Bayesian methodology, namely, the regularized maximum likelihood following Ljung (1999). Benes and others have produced a template for such a model- based, multivariate filtering technique which avoids some of shortcomings of previous approaches. It also provides a common yet flexible framework in which different countries can be more easily compared. The multivariate model incor- porates relevant empirical relationships between actual and potential GDP, including unemployment, core inflation, and capacity utilization. The model is applied here to quarterly data for Germany from 1991:Q1 to 2010:Q4. For comparison purposes, some estimates for the United States are also shown. The model is in many ways a standard, reduced-form macroeconomic model built around three gaps: an output gap, an unemployment gap, and a capacity utilization gap. These gaps are pinned down by a number of identifying equations, including an inflation equation that relates inflation to the output gap through a Taylor-rule relationship; an unemployment equation that estimates an Okun’s law relationship; and a capacity utilization equation to exploit the information that capacity utilization rates can provide in the estimation of the aggregate output gap. Thus, to run the model, five standard data series are needed (Table 2.1). The appendix contains a more detailed summary of the main model features, as does Benes and others (2010). The methodology requires taking a stance on prior beliefs regarding a number of variables. Consistent with past economic patterns, a key assumption fed into the model’s estimation is that demand shocks are the primary source of real GDP fluctuations in Germany. As discussed above, Germany’s economy has been increasingly (export) demand driven, and especially so during the most recent crisis. An accurate understanding of potential GDP in Germany must take into consideration these facts.

In the model, the importance of demand versus supply shocks is implemented GU Y through the (relative) priors on the standard deviations of εt (demand) and εt

©International Monetary Fund. Not for Redistribution 42 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

Multivariate model (demand) Multivariate model (supply) 7.0 Hodrick-Prescott filter Moving average (12 quarters) 5.0

3.0

1.0

−1.0

−3.0 −5.0 −7.0 1997:Q1 1997:Q4 1998:Q3 1999:Q2 2000:Q1 2000:Q4 2001:Q3 2002:Q2 2003:Q1 2003:Q4 2004:Q3 2005:Q2 2006:Q1 2006:Q4 2007:Q3 2008:Q2 2009:Q1 2009:Q4 2010:Q3 Figure 2.5 Output Gaps Based on Alternative Methodologies (Percent of potential GDP) Sources: IMF staff estimates.

(supply) (see the appendix regarding where these variables enter the model). For example, a prior belief that supply is more volatile than demand would lead the model to assign much of the observed volatility of real GDP to potential GDP fluctuations. Put differently, actual and estimated potential GDP would move in sync, and the output gap would be less volatile. Conversely, if most GDP volatility is attributed to demand shocks, then potential GDP (supply) would remain more stable, resulting in a more volatile output gap. Both the model assumptions and the choice of methodology matter, affecting the results that are obtained. Alternative methods of calculating potential GDP (and the resulting output gaps) include simple filters that take a moving average of actual GDP and methods based on the Hodrick-Prescott (HP) approach (see Hodrick and Prescott, 1981), which estimates a statistically smoothed series.8 Also, within the framework used here, as discussed, different priors on the distribution of shocks to GDP (demand versus supply shocks) matter. Notably, with the exception of the fairly crude moving-average approach, all approaches deliver fairly similar results historically—they do differ, however, in their output gap dynamics during the current crisis (Figure 2.5). That is, all atheoretical approaches assign most of the variation in actual GDP to supply variations (i.e., variations in potential GDP) and thus exhibit a similar pattern as that resulting from the multivariate approach with supply-sided priors. Given the nature of the shock that Germany experienced during this crisis, these other approaches are likely to be misleading. That is, not utilizing information on the underlying economics, including especially the nature of the shock, provides a very different view of the evolution of potential GDP than otherwise.

8 That is, the HP-filter does not use economic information in its estimation. It also suffers from end- of-sample measurement problems.

©International Monetary Fund. Not for Redistribution Schindler 43

France Germany Japan 12.0 United Kingdom United States

10.0

8.0

6.0

4.0

2.0

0.0 2001:Q1 2001:Q3 2002:Q1 2002:Q3 2003:Q1 2003:Q3 2004:Q1 2004:Q3 2005:Q1 2005:Q3 2006:Q1 2006:Q3 2007:Q1 2007:Q3 2008:Q1 2008:Q3 2009:Q1 2009:Q3 2010:Q1 2010:Q3 Figure 2.6 Unemployment Rates (Percent) Sources: IMF; World Economic Outlook; and IMF staff estimates.

The projected potential growth patterns imply virtually no medium-term loss in potential German GDP. The mild drop in potential growth in Germany, combined with potential growth slightly above the precrisis trend growth in the medium term, suggests that Germany may not suffer any medium-term losses in its level of potential output. According to the model projections, by 2016 potential GDP should converge with the level of potential output that would have been reached had precrisis growth rates prevailed throughout the projection period.9 By contrast, the sharper drop in potential growth in the United States, combined with model-based potential growth projections that fall short of precrisis rates, implies that the United States may suffer a substantial permanent loss in potential GDP. Even at end-2016, U.S. potential GDP is estimated to be nearly 10 percent below the level based on precrisis trends. The mild potential GDP dynamics in Germany have been mirrored by a similarly mild labor market response (Figure 2.6). A shock that leaves medium- term potential GDP relatively unharmed in both its level and its growth reduces the incentives for firms to shed employment, particularly when such labor hoarding is supported by additional policy measures. The moderate and only short- lived uptick in German unemployment partly reflects a shock that was temporary and demand-sided. In turn, the low unemployment impact reinforces the positive potential GDP dynamics: keeping unemployment down helps avoid many of the

9 Using a multivariate state-space model, El-Shagi (2011) reaches a very similar conclusion for Germany and finds “that the crisis mostly opened the output gap and did not reduce potential GDP (as suggested by other models)” (p. 737).

©International Monetary Fund. Not for Redistribution 44 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

5 Euro Area Germany United States 3

−1

−3

−5

−7 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Figure 2.7 Output Gaps (Percent of potential GDP) Source: IMF, World Economic Outlook.

adverse short-term effects of unemployment on potential output.10 Thus, potential output and labor market dynamics form a consistent equilibrium. By contrast, the United States appears to have embarked on a different equilibrium path, where a more supply-side shock has led to a larger impact on potential GDP and reduced incentives for firms to retain employment. The different labor market experiences observed in Germany and the United States therefore provide a consistent mirror image to their estimated potential GDP dynamics. The robust recoveries in actual GDP and capacity utilization imply an output gap that was projected to be closed by end-2011 at the time of estimation11 (Figure 2.7). Given the fairly stable path of potential output, the drop in actual GDP during the crisis led to a large negative output gap; conversely, the growth recovery implies a shrinking of the output gap, closing in 2012 and remaining near zero from then onwards. Capacity utilization follows a similar path (Figure 2.8). The broad output gap dynamics are similar across most sectors, with gaps in most sectors still negative, though close to closed in some (Figure 2.9). With a still negative output gap, inflationary pressures remain moderate (Figure 2.10). Based purely on the projected output gap dynamics, core inflation was projected to increase only moderately, to about 1.3 percent in 2011, from

10 Job creation is a more sluggish process than job destruction, suggesting some persistence in unemployment even after a (temporary) shock has subsided; also, unemployment is often associated with losses in human capital associated with unemployment spells. Both factors imply a reduction in effective labor supply and thus potential output. In “What Does The Crisis Tell Us About the German Labor Market?”—Chapter 4 in this volume—I examine in more detail some of the factors behind the moderate labor market response. In addition to the potential GDP dynamics, past reforms, including the Hartz IV in 2005, were likely contributors. 11 The work underlying this chapter was completed in early 2011.

©International Monetary Fund. Not for Redistribution Schindler 45

4 Relative to historical average (1991:Q1–2008:Q1) Relative to pre-crisis peak 2

−2

−4

−6

−8 Consumer Intermediate Capital goods Consumer durables goods nondurables

Figure 2.8 Capacity Utilization Gaps (Percent) Sources: Federal Statistical Office; Deutsche Bundesbank; and IMF staff estimates.

25 Agriculture Manufacturing Total Value Added Services 20 15 10 5 0 −5 −10 −15 −20 1991:Q1 1992:Q1 1993:Q1 1994:Q1 1995:Q1 1996:Q1 1997:Q1 1998:Q1 1999:Q1 2000:Q1 2001:Q1 2002:Q1 2003:Q1 2004:Q1 2005:Q1 2006:Q1 2007:Q1 2008:Q1 2009:Q1 2010:Q1 Figure 2.9 Sectoral Output Gaps (Percent) Sources: Federal Statistical Office; Deutsche Bundesbank; and IMF staff estimates. Note: Measured as log-differences between actual value added and an estimated linear time trend. about 0.8 percent in 2010.12 Second-round effects on wages have also so far remained moderate, not least reflecting a still-positive unemployment gap (i.e., actual unemployment above equilibrium). And lastly, aggregate capacity utilization is set to remain substantially below precrisis levels. However, continued and sharper commodity price pressures do represent an upside risk to core inflation. Based on the model estimated here, the crisis impact on potential GDP is more limited in Germany than in the United States. From a historical (post-unification) average of about 1.3 percent, German potential growth declined moderately to

12 An increase in the output gap by one percentage point is associated with about a 2/5 percentage point increase in inflation.

©International Monetary Fund. Not for Redistribution 46 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

Output gap Core inflation (year on year) 8 Capacity utilization gap Unemployment gap 6 4 2 0 −2 −4 −6 −8 −10 −12 1995:Q3 1996:Q3 1997:Q3 1998:Q3 1999:Q3 2000:Q3 2001:Q3 2002:Q3 2003:Q3 2004:Q3 2005:Q3 2006:Q3 2007:Q3 2008:Q3 2009:Q3 2010:Q3 2011:Q3 2012:Q3 2013:Q3 2014:Q3 2015:Q3 2016:Q3 Figure 2.10 Germany: Gap Measures and Inflation (Percent) Sources: IMF, World Economic Outlook; Deutsche Bundesbank; and IMF staff estimates.

about 1.0 percent in 2009. However, in the medium term, pulled up by a strong growth recovery, potential growth is expected to move slightly above historical rates, peaking at about 1.7 percent in 2013 before eventually converging back to its long-term rate of 1.3 percent by 2016. By contrast, the U.S. economy is much more closed; and the crisis there originated from the financial and real estate sectors, suggesting a domestic supply shock. Thus, U.S. potential growth declined more dramatically, to about 1.5 percent in 2009 from over 2.5 percent historically. In the medium term, potential growth in the United States is projected to pick up to about 2¼ percent by 2016 (Figures 2.11 and 2.12).

3 Germany USA

0 2003 2005 2007 2009 2011 2013 2015

Figure 2.11 Germany and the U.S.: Potential Growth (Percent) Sources: IMF, World Economic Outlook; and IMF staff estimates.

©International Monetary Fund. Not for Redistribution Schindler 47

125 Germany Germany pre-crisis trend 120 United States 115 United States pre-crisis trend 110

105

100

90 2003 2005 2007 2009 2011 2013 2015

Figure 2.12 Germany and the U.S.: Potential GDP Relative to Pre-crisis Trends (2007=100) Sources: IMF, World Economic Outlook; and IMF staff estimates.

GERMANY’S GROWTH SOURCES THROUGH A GROWTH ACCOUNTING LENS Though the crisis did little damage to Germany’s growth potential, that potential remains low in cross-country comparisons. Germany’s economic potential has sur- vived the crisis remarkably intact. In contrast, U.S. potential growth may be reduced by a significant fraction of its historical rates. Yet, even at reduced growth rates, U.S. potential growth is still nearly twice Germany’s long-term potential of 1.3 percent. What accounts for Germany’s low potential growth rate? The growth accounting approach, which decomposes GDP growth into that arising from TFP growth, changes in the capital stock and employment growth, can yield valuable insights (See Table 2.2). Building on the estimates by Bosworth and Collins (2003) and applying their approach to quarterly data suggests that: • short-term fluctuations in output are largely driven by fluctuations in TFP (the residual); and • longer-term trends in German growth prior to the crisis have been driven by both TFP growth and capital accumulation. During 1991–2010, GDP grew at an average 1.3 percent per year, of which 0.5 percentage points were due to TFP growth, 0.6 percentage points due to growth in the capital stock, and 0.1 percentage points due to employment growth.13

13 Different labor measures can be used, but do not affect the estimates substantially. When labor input is measured as the labor force rather than employment, its contribution is higher at 0.2 percent, but TFP growth remains at about 0.5 percent annually. Estimated average TFP growth rises to a slightly higher value of about 0.75 percent annually when labor is measured as total hours, reflecting a secular decline in total hours worked.

©International Monetary Fund. Not for Redistribution 48 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

TABLE 2.2 Decomposing growth (percent, quarter-on-quarter, annualized) GDP TFP Capital Labor 2000:Q1–2005:Q4 0.9 0.4 0.5 0.1 2006:Q1–2008:Q2 4.7 2.2 0.9 0.2 2008:Q3–2009:Q1 –5.5 –6.0 0.5 0.0 2009:Q2–2010:Q2 3.9 3.3 0.3 0.3 2010:Q3–2010:Q4 2.1 1.1 0.4 0.4 Average 1991–2010 1.3 0.1 0.6 0.1

Source: IMF staff estimates. Note: TFP: total factor productivity.

6 5 Staff estimates EU Klems Conference Board 4 GGDC OECD 3 2 1 0 −1 −2 −3 −4

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

Figure 2.13 Germany: Total Factor Productivity Growth Estimates (Percent). Sources: IMF staff estimates. Note: GGDC: Groningen Growth and Development Centre; Klems: growth and productivity accounts.

These estimates are broadly consistent with a variety of other estimates. The EU KLEMS project, the Conference Board, the Groningen Growth and Development Centre (GGDC) and the OECD are the main alternative sources of TFP estimates. While these measures differ somewhat in detail, they provide broadly similar estimates and are highly correlated.14 In particular, they all point to a secular downward trend in TFP growth that only briefly picked up in the run-up to the crisis. Consistent with our estimates of a limited impact of the crisis on potential GDP, the post-crisis TFP estimates exhibit a sharp rebound15 (Figure 2.13).

14 Differences in a number of variables and parameters can affect TFP estimates, including capital and labor shares and capital stock estimates, with the latter especially sensitive to different assumptions (such as the initial stock and the rate of depreciation). 15 According to staff estimates, TFP declined by more than 5 percent in 2009, but grew by nearly 3 percent in 2010.

©International Monetary Fund. Not for Redistribution Schindler 49

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

Italy Japan Spain Austria France Canada Sweden AustraliaBelgium Denmark Netherlands United States New Zealand United Kingdom Germany (OECD) Germany (IMF staff)

Figure 2.14 Total Factor Productivity Growth (Average annual growth, 1992–2009, percent) Source: Organisation for Economic Co-operation and Development; and IMF staff estimates.

In the medium term, further labor market improvements would temporarily sustain a higher potential growth rate. The downward trend in the unemployment rate that started around 2005, interrupted only briefly by the crisis, may reflect ongoing dividends from the Hartz IV labor market reforms. If so, the continued transition to a lower structural unemployment rate could temporarily support potential growth above its long-term rate of 1.3 percent.16 In addition, higher labor market participation rates and increased immigration could help ease the impact of demographic pressures and further support employment growth (Figure 2.14). However, a longer-lasting increase in the growth potential can only be achieved through higher TFP growth or faster capital accumulation. Germany’s TFP growth lags that of many of its peers (see the discussion in section B). While the United States experienced annual TFP growth of about 1.2 percent annually during 1992–2009, German productivity expanded by only 0.4 percent a year (0.9 percent according to OECD data). (Figure 2.15) Sectoral data suggest that the productivity shortfall reflects lags in investment in information and communication technology (ICT) capital and related innovations, especially in private services (see Chapter 3, by Helene Poirson, in this volume). Higher investment rates could boost both capital accumulation and the adoption and development of new technologies.17

16 See also Schindler (2012) for an argument that the Hartz reforms (which lowered social transfer payments and deregulated temporary employment) had a long-lasting effect on job creation that is still being felt. In particular, in that view, the secular downward trend since 2005 represents the transition to a lower steady-state unemployment rate. 17 High-skilled immigration, in addition to expanding the labor supply, may also contribute to TFP growth. See West’s (2011) case for revamping the U.S. immigration system.

©International Monetary Fund. Not for Redistribution 50 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

120 USA UK Germany Japan Euro Area 115

110

105

100

80 2000 2002 2004 2006 2008 2010 2012 2014 2016

Figure 2.15 Real Per Capita Potential GDP (2007=100) Sources: IMF, World Economic Outlook; and IMF staff estimates.

The importance of productivity improvements is especially pertinent in a long-term context. Given Germany’s declining population, although its potential GDP is set to rise more strongly than elsewhere in per-capita terms, standard growth theory suggests that long-run growth in living standards is driven by productivity growth. Thus, policies that support TFP to maintain the medium-term improvement in living standards are also crucial in the long term.

CONCLUSION Germany has mastered the crisis well, but long-term challenges remain. Reflecting the nature of the shock and the structure of its economy, Germany has emerged from the crisis remarkably unblemished. However, it is set to return to its previous trajectory of only moderate growth. Raising that growth rate will require policy measures on many fronts, including higher labor-force participation, higher investment rates, and—especially—stronger productivity growth. More recent developments in Germany’s growth recovery also pose some downside risks to the view taken here. The medium-term growth prospects have fallen short of most observers’ expectations, inducing downward revisions of many analysts’ growth projections, including those by the IMF. The absence of a permanent loss in the level of potential output depended on above-average potential growth in the near term, to compensate for the (mild) potential growth slowdown during the crisis. To the extent that lower-than-expected growth also reduces potential growth, potential output might not fully recover to its precrisis trend. Such downside risks underscore the need for further reforms to accelerate long-term potential growth in Germany.

©International Monetary Fund. Not for Redistribution Schindler 51

REFERENCES Baghli, Mustapha, Christophe Cahn, and Jean-Pierre Villetelle, 2006, “Estimating Potential Output with a Production Function for France, Germany and Italy,” Banque de France Working Paper No. 146 (Paris: Banque de France). Benes, Jaromir, Kevin Clinton, Roberto Garcia-Saltos, Marianna Johnson, Douglas Laxton, Petar Manchev, and Troy Matheson, 2010, “Estimating Potential Output with a Multivariate Filter,” IMF Working Paper No. 10/285 (Washington, DC: International Monetary Fund). Billmeier, Andreas, 2004, “Ghostbusting: Which Output Gap Measure Really Matters?,” IMF Working Paper No. 04/146 (Washington, DC: International Monetary Fund). Bosworth, Barry, and Susan M. Collins, 2003, “The Empirics of Growth: An Update” (Washington, DC: Brookings Institution). Cahn, Christophe, and Arthur Saint-Guilhem, 2007, “Potential Output Growth in Several Industrialised Countries: A Comparison,” ECB Working Paper No. 828 (Frankfurt am Main: European Central Bank). De Masi, Paula, 1997, “IMF Estimates of Potential Output: Theory and Practice,” IMF Working Paper No. 97/177 (Washington, DC: International Monetary Fund). Deutsche Bank Research, 2011, “Outlook 2011: German Growth Remains Robust,” Germany: Current Issues, February 14, 2011, downloaded from http://www.dbresearch.com. El-Shagi, Makram, 2011, “Did the Crisis Affect Potential Output?,” Applied Economics Letters, Vol. 18, pp. 735−38. Furceri, Davide, and Annabelle Mourougane, 2009, “The Effect of Financial Crises on Potential Output: New Empirical Evidence from OECD Countries,” OECD Economics Department Working Paper No. 699 (Paris: Organisation for Economic Co-operation and Development). Gordon, Robert J., 2008, “The Slowest Potential Output Growth in U.S. History: Measurement and Interpretation,” Paper prepared for presentation at the CSIP Symposium, “The Outlook for Future Productivity Growth” (San Francisco: Federal Reserve Bank of San Francisco). Hodrick, Robert J., and Prescott, Edward C., 1981, “Postwar U. S. Business Cycles: An Empirical Investigation.” Discussion Paper No. 451 (Minneapolis: University of Minnesota). Horn, Gustav, Camille Logeay and Silke Tober, 2007, “Estimating Germany’s Potential Output,” IMK Working Paper No. 2/2007 (Washington, DC: International Monetary Fund). Ljung, Lennart, 1999, System Identification: Theory for the User (Princeton, New Jersey: Prentice- Hall). Mankiw, N. Gregory, 2002, Macroeconomics, 5th edition (New York: Worth Publishers). Sachverständigenrat (Sachverständigenrat zur Begutachtung der gesamtwirtschaftlichen Entwicklung), 2011, “Herausforderungen des demografischen Wandels,” Expertise im Auftrag der Bundesregierung, May 2011. Vitek, Francis, 2010, “Output and Unemployment Dynamics During the Great Recession: A Panel Unobserved Components Analysis,” IMF Working Paper No. 10/185 (Washington, DC: International Monetary Fund). West, Darrell M., 2011, “Creating a ‘Brain Gain’ for U.S. Employers: The Role of Immigration,” Brookings Policy Brief No. 178 (Washington, DC: The Brookings Institution).

©International Monetary Fund. Not for Redistribution 52 The Crisis’s Impact on Potential Growth in Germany: The Nature of the Shock Matters

APPENDIX This appendix briefly summarizes the key features of the model. A more detailed description is provided in Benes and others (2010). The model is built around three gaps—the output gap (y), the unemployment gap (u), and the capacity utilization gap (c)—and three identifying equations: The inflation equation relates the level and the change of the output gap to core inflation:

πΩ=π +β + − +επ4 44tt−−11yyy t() ttt.

The dynamic Okun’s law defines the relationship between the current unemployment rate and the output gap. Based on Okun’s law, an unemployment equation links the unemployment gap to the output gap:

=++εu uxuxytt11− 2 tt.

Finally, the model also relies on a capacity utilization equation, on the assumption that capacity utilization contains important information that can help improve the potential output and output gap estimates. The equation takes the following form:

=++εc cttk11c − k 2y tt.

Given the three identifying equations, the equilibrium variables are assumed to evolve dynamically as follows. A stochastic process including transitory (level) shocks and more persistent shocks guides the evolution of equilibrium unemploy- U ment ()t (the NAIRU equation): ωλ USSU UU=+−tt−−11 G y− −() U −+ε U ttt1001 100 t

U Persistent shocks to the NAIRU (Gt ) follow an autoregressive process:

U GGUUG=−α(1 ) +ε (1) ttt−1

And potential output ( Yt ) is modeled to be a function of the underlying trend Y growth rate of potential output (Gt ) and changes in the NAIRU. Specifically:

YY YY=−θ−−−θ−t −1 ()( UUtt−−−11201 )( U t U t )/19 + G /4 +ε (2) t tt where θ is the labor share in a Cobb-Douglas production function. This specification allows for short- and medium-term growth of potential to differ from trend Y growth. Note that Gt is not constant, but follows serially correlated deviations

©International Monetary Fund. Not for Redistribution Schindler 53

Y (long waves) from the steady-state growth rate GSS . Similar dynamic equations are specified for equilibrium capacity utilization. The full model is estimated by regularized maximum likelihood (Ljung, 1999), a Bayesian methodology. This method requires the user to define prior distributions of the parameters. While this can improve the estimation procedure by preventing parameters from wandering into nonsensical regions, the choice of priors has also non-negligible implications for the final estimates as the data are uninformative about some parameters. The choice of priors matters also in the German case. In addition to the prior distributions of parameters, the analyst has to provide values for θ and the steady-state (long-run) unemployment rate (U SS ) and poten- Y tial GDP growth rates (GSS ), which were set to 0.55, 6.7 percent and 1.3 percent, SS Y respectively. While values especially for U and GSS matter conceptually, as the (endogenous) estimates converge to these (exogenously given) values in the long term, from a practical point of view, the dynamics over the time horizon of interest are relatively little affected by the choice of the steady-state values.

©International Monetary Fund. Not for Redistribution This page intentionally left blank

©International Monetary Fund. Not for Redistribution CHAPTER 3 German Productivity Growth: An Industry Perspective

HÉLÈNE POIRSON

Germany had the advantage of a catch-up phase through the early 1970s, but thereafter, and especially since the mid-1990s, productivity growth has been relatively slow. This aggregate picture masks considerable intersectoral differences. This chapter reports that while Germany has maintained an enduring strength in its traditional manufacturing competencies, it has lagged in services delivery and, perhaps more importantly, in the so-called knowledge economy—both in the production of information and communication technology (ICT) goods and in the use of ICT to raise productivity in private services provision. These “new” areas have been key engines of productivity growth since the mid-1990s in the United States and some other advanced countries. An uptick in German services productivity growth occurred from 2005 to 2007. This is indicative of the scope and potential for progress, but it could have been a cyclical phenomenon. To foster more efficient production and widespread use of ICT in services delivery will require complementary measures: (a) further development of venture capital and private equity markets (backed by a more efficient insolvency process); (b) increased commercial use of intellectual property rights held by university and research institutions; (c) removal of uncertainties regarding tax treatment; and (d) further European integration in services provision.

INTRODUCTION In the years after World War II, Germany increased its productivity at a fast clip, thereby rapidly shrinking the United States’ lead in income levels. However, this productivity catch-up was interrupted in the mid-1990s. Between 1995 and 2007, Germany’s private economy experienced annual productivity growth of 1.7 percent, slightly above the European average but lagging behind the United States, the United Kingdom, Sweden, Austria, and Finland (Inklaar, Timmer, and van Ark, 2008, Table 1, page 142; and Molagoda and Perez, 2011, Table V.1, page 70).

For their constructive comments, the author wishes to thank Fabian Bornhorst, Malte Hübner, Anna Ivanova, Ashoka Mody, Martin Schindler, and participants in the Germany in an Interconnected World Conference at the BMF in Berlin (May 2011). 55

©International Monetary Fund. Not for Redistribution 56 German Productivity Growth: An Industry Perspective

BOX 3.1 The Productivity Resurgence in the United States

Jorgenson, Ho, and Stiroh (2008) argue that the sources of U.S. productivity growth have changed twice since 1995. From 1995 to 2000, productivity growth was led by the ICT- producing sectors and investment in them. Since 2000, the sources of productivity growth have shifted to sectors that were the most intensive users of information technology. Triplett and Bosworth (2006) highlight the role of the services sector. They find that the post-1995 ICT investment boom contributed both directly to higher U.S. productivity growth in the services industries via capital deepening and indirectly via higher TFP growth. Their results suggest that in the United States, both labor productivity in the services sector and TFP growth rates more than doubled after 1995, and the services sector became the source of economic growth. In a subsequent paper, the same two authors extend the industry-based approach to consider the post-2000 period and find that the services sector again accounts for the post- 2000 rise in U.S. productivity (Bosworth and Triplett, 2007). Both the 1995–2000 increase in productivity growth and the more recent uptick were driven by strong investment that increased U.S. labor productivity through capital deepening and by accelerating productivity (particularly TFP) in services.

This aggregate performance masks considerable sectoral differences—both within Germany and in Germany’s performance relative to other industrialized economies. While Germany’s productivity growth in manufacturing industries was above the EU-15 average from 1995 to 2005, its productivity growth in the information and communication technology (ICT) industries, while strong, has been lagging that in other advanced countries (U.S, France, Netherlands, and Sweden), and productivity in the private services sector has been weak and below that in both the United States and the EU-15. Within Germany, the deceleration in overall productivity growth in the private economy—from 2.4 percent in the 1980s to 1.9 percent during 1995–2000 and dropping further to 1.2 percent in 2000–05—appears to be largely due to the slowing growth of productivity in private services (The Conference Board, 2009). This contrasts with the experience of the United States and selected other countries over the same period, when productivity accelerated from 1995 to 2005 on the back of ICT’s widespread and quick deployment in many sectors of the economy, especially the traditionally less productive service sectors (Box 3.1).1 Germany’s experience mirrors that of Europe as a whole. An industry-based analysis applied to European countries finds that while the EU-15’s manufacturing and utilities performed in line with the United States during 1995–2005, local services—not just financial services, but more broadly distribution and business services—accounted for the bulk of the productivity growth difference with the United States during that period (van Ark, O’Mahony, and Timmer, 2008,

1 Jorgenson, Ho, and Stiroh (2004) first highlighted the role of the production and use of ICT in accounting for productivity gains in the case of the United States.

©International Monetary Fund. Not for Redistribution Poirson 57 and Molagoda and Perez, 2011).2 Unlike in the United States, where total factor productivity (TFP) growth in the services sector accelerated after 1995, in Europe it declined (especially in distribution, finance, and business services). Several studies have argued that low TFP growth in the service sectors in EU countries in turn reflects the delay in investing in ICT assets and implementing the related new technologies and processes (McKinsey Global Institute, 2010). These delays have been attributed, among other factors, to the relatively less attractive conditions in Europe for venture capital and other market-based sources of financing for high-growth, high-risk projects (Box 3.2).3 In addition to the availability of long-term financing, traditional explanations for Europe’s lagging productivity growth also emphasize policy and institutional factors such as strict regulations in product and labor markets.4 But these seem less relevant in the case of Germany, which is one of the most deregulated advanced countries in retail trade, according to OECD indicators, and where some reforms (e.g., transport services deregulation) constitute an example of best practice in Europe. It is true, however, that despite a recent improvement Germany remains more heavily regulated in professional services than most other advanced countries. This study applies the industry-based approach to Germany during the 1980–2007 period to help shed light on the historical sources of growth and productivity trends. We use the best available data on cross-country comparisons of the sources of productivity growth at the industry level from the EU KLEMS database. This enables us to document a pick-up in labor productivity growth in the private economy in Germany since 2005, led by higher TFP growth. This is indicative of the scope and potential for progress. However, it remains too early to conclude that a structural shift has taken place. The fact that TFP is measured as a residual and the finding that the TFP growth acceleration was not accompanied by a boom in ICT investment—unlike the U.S. experience in the late 1990s and early 2000s—suggests that the recent improvement may be short-lived and driven by cyclical rather than structural factors.

2 An argument is often made that productivity differentials, particularly in the services industries, are biased or illusionary because of differences in data across countries. The results in Inklaar, Timmer, and van Ark (2006), however, suggest that the productivity gap findings are robust to the use of various productivity measurement models. A related argument is that the strong post-1995 productivity growth in the U.S. services sector was illusionary because it was the result of an unsustainable boom in consumer expenditure and household debt. While scale effects from increased demand may have played a role in productivity improvements—notably for the distribution sector—the fact that the acceleration in retail output and productivity occurred mainly during the late nineties and early 2000s, whereas the rise in household debt occurred later, from 2003 to 2007, makes it difficult to attribute all the productivity improvements to the credit boom (van Ark, 2010). 3 Allard and Everaert (2010) argue that measures to develop capital markets further in Europe will not only result in the establishment of more attractive conditions for venture capital but also create room for banks to focus more on supporting smaller firms—in relatively large numbers in the euro area; these smaller firms are at a higher risk of being constrained in financing but at the same time are key for innovation. 4 See Allard and Everaert (2010) for a discussion of the role of labor and service market reforms in lifting euro area long-term growth.

©International Monetary Fund. Not for Redistribution 58 German Productivity Growth: An Industry Perspective

BOX 3.2 The Slowdown in Europe’s TFP Growth

There is no shortage of attempts to explain the post-1995 weak European TFP growth performance. Several studies have documented that Europe’s lag in adopting ICT technologies in the service sectors and shifting to a “knowledge economy” is a key factor behind Europe’s poor TFP performance (see for example van Ark, O’Mahony, and Timmer, 2008). Empirical research has confirmed the role of “new economy” factors, such as human capital and ICT investments, in explaining subsequent TFP growth rates. For example, Molagoda and Perez (2011) find that a significant impact is made by human capital levels (proxied by the share of high skilled labor to overall labor) and ICT capital intensity (measured by the ratio of ICT capital to non-ICT capital) on TFP growth differences across industries and countries (see also Nahuys and van der Wiel, 2005, and EU ICT Task Force, 2006). Other studies argue that policy/institutional factors (e.g., rigidities in product, labor, and financial markets) have reduced incentives to shift rapidly to ICT and to adjust production processes accordingly. Using micro establishment level data from the U.S. and Germany, the results in Bartelsman and others (2010) suggest that the degree of market experimentation by firms (i.e., the degree to which firms experiment with different ways of conducting business) is lower in Germany compared to the United States, among both young businesses and businesses actively changing their technology. The authors conjecture that this result could be related to the financial system being more market-based in the United States than in Europe, which possibly lowers risk aversion to project financing and creates greater financing possibilities for entrepreneurs with small and innovative projects.a The effect of strict employment protection legislation (EPL) is less clear-cut and largely depends on the institutional system in which firms operate and the type of technology used in the sector.b

a Voss and Müller (2009) similarly find that financing aspects are a key limiting burden for young German start-ups. Venture capital is hard to access and concentrates on ICT, medical research, medical appli- ances, and biotech. b Bartelsman and others (2010) investigate the impact of EPL on ICT adoption. They find evidence that both the share of employment and productivity levels in ICT-intensive sectors are relatively lower in high-protection EU countries, suggesting that EPL slows the adoption of new ICT. However, they do not investigate the direct impact of EPL on TFP growth and innovation.

Using the historical experience of the United States as a productivity leader from 1995–2004 as a benchmark, our conclusions highlight the importance of raising TFP growth more durably through innovation policy and greater incentives for the use of ICT in the service sectors5: • Over 1995–2004, Germany’s lower TFP growth was the single most important factor driving German-U.S. productivity growth differences in the private economy, contributing almost three-quarters of the productivity growth differential.

5 The importance and cross-cutting nature of ICT is reflected in Germany’s high-tech strategy and, more broadly, in Europe’s 2020 growth agenda, which also notes the link between ICT usage and services’ productivity.

©International Monetary Fund. Not for Redistribution Poirson 59

• Germany’s lower investment in ICT assets was also important, explaining over a third of the productivity gap with the United States during that period. • At the industry-level, the productivity growth differences are driven by TFP and are particularly large for service industries, especially for distribution, finance, and business services. • Productivity lags in the service sectors in turn appear to reflect Germany’s delay in shifting to a knowledge economy, as indicated by a lower share of ICT in total research and development expenditure, lower internet penetration (relative to the United States), and lower ICT readiness scores relative to some other advanced countries. Notably, Germany lags several other OECD countries in some key areas of ICT infrastructure (e.g., availability of secure servers) and financing for high-risk, high-innovation potential projects. The purpose of this chapter is to provide background on the question of why Germany has not enjoyed higher productivity growth, despite several favorable competitiveness indicators, including levels of innovation and spending on research and development above the EU-15 average (Figure 3.1).6 It does so by providing essential facts and placing Germany’s experience in international perspective. The next section provides facts on aggregate output and growth patterns in Germany relative to the United States (considered as the productivity benchmark) during 1950–2009. A further section does the same for a number of

Finland Sweden Netherlands Germany Denmark Japan Austria Belgium United States Norway France United Kingdom Italy Spain

0 100 200 300 400 500 600

Figure 3.1 Selected Countries: Patent Applications (Number of applications per million population, 2010) Sources: European Patent Office and Conference Board.

6 See McKinsey Global Institute (2010). Recent structural reforms to support competitiveness include the 2008 business tax reform, the Hartz IV reforms, and deregulation in transport services.

©International Monetary Fund. Not for Redistribution 60 German Productivity Growth: An Industry Perspective

industry-level sectors, including goods, ICT, and private services, extending earlier work by van Ark, O’Mahony, and Timmer (2008) through 2007. Within the services sector, we separately analyze the drivers of growth and German-U.S. productivity differentials in distribution services, finance and business services, and personal services. The section also documents Germany’s relative progress in adopting a knowledge economy, as measured by a number of variables and business survey indicators usually viewed as related to higher ICT investments and a favorable environment for innovation.

GERMAN AND UNITED STATES PRODUCTIVITY: STYLIZED FACTS Germany’s postwar productivity catch-up with the United States was interrupted in the mid-1990s. During the period 1950–1995, labor productivity grew at a relatively fast clip, and Germany achieved parity with the United States by 1995. In 1995, however, German productivity growth started lagging the United States, and by 2009 its levels of hourly output were almost 10 percent below U.S. levels (Figure 3.2 and Table 3.1). During 1995–2004, the productivity gap with the United States widened despite higher labor input in the latter country, where the productivity resurgence in the mid-1990s was accompanied by significant job creation. Instead, the productivity gap reflects both slower capital accumulation and slower TFP growth in Germany. Germany’s overall slower capital accumulation contributed almost three-quarters (74 percent) of the total productivity gap during 1995–2004. In particular, its lower investment in ICT assets—including computing equipment,

120 GDP per hour 100

80 GDP per capita

0 1950 1960 1970 1980 1990 2000

Figure 3.2 Total Economy GDP per Hour Worked and GDP per Capita (Percent of U.S. levels) Sources: The Conference Board Total Economy Database, September 2010; Timmer, Ypma, and van Ark (2003); and IMF staff estimates.

©International Monetary Fund. Not for Redistribution Poirson 61

TABLE 3.1 Levels of Labor Productivity and Factor Inputs (Percent of U.S. levels) 1950 1973 1995 2004 2007 2008 2009 GDP per capitaa 43.6 82.5 85.1 77.7 78.9 80.7 79.5 Hours worked per capita 124.8 104.4 88.3 84.3 83.7 83.8 82.7 GDP per hour workeda 34.1 70.2 100.3 94.0 95.9 95.2 90.6 Capital input per hour workedb 83.9 107.9 102.7 n.a. n.a. n.a.

Sources: The Conference Board Total Economy Database, September 2010; Timmer, Ypma, and van Ark (2003); and IMF staff estimates. a Output levels are converted by GDP purchasing power parities for 2009 using the Elteto-Koves-Szulk (EKS) method. b Measured as gross fixed capital stock per hour worked. Entries for 1973 refer to 1980. communication equipment, and software—was the single biggest factor explaining the productivity gap during that period, contributing over half of it (0.6 percent out of an annual productivity gap of 1.1 percent with the United States). The contribution of non-ICT capital growth and TFP growth to the productivity growth differential with the United States were also important: the former contributed 0.3 percent (22 percent of total) and the latter 0.2 percent (15 percent of total) to the productivity gap. Lower growth in labor skills in Germany relative to the United States explained the remainder (0.1 percent, or about 10 percent of the total). When focusing on developments over time in Germany and the United States, a deceleration in TFP growth explains the bulk of Germany’s weak performance since the mid-1990s.7 German labor productivity growth slowed by 1 percentage point between the 1990–1995 period and the 1995–2004 period, mainly reflecting a reduction in TFP growth by 0.6 percentage points between the two periods (Table 3.2). In other words, slower TFP growth accounts for 60 percent of the deceleration in annual labor productivity growth during 1995–2004, with lower investment and (to a much lesser extent) slower growth in labor quality explaining the rest.8 Investment in ICT assets in Germany increased only moderately between 1995–2004 and 1990–1995, not enough to offset the decline in TFP and investment in non-ICT assets. In contrast, higher TFP growth and investment—especially in ICT assets—contributed in almost equal parts to the U.S. revival in the mid-1990s. Specifically, the 1 percentage point increase in U.S. labor productivity growth between 1995–2004 and 1990–1995 reflects a 0.5 percent increase in the contribution of capital accumulation (the bulk of which is from investment in ICT assets), accompanied by a 0.4 percent acceleration in TFP growth. This simultaneous acceleration in ICT investments and TFP has been linked to the emergence of a so-called “knowledge economy” in the United States and other advanced countries during the mid-1990s, as discussed in

7 TFP is computed as a residual, thus it also includes measurement errors and the effects from unmeasured outputs and inputs, such as research and development and other intangible improvements, including organizational improvements. 8 Lower investment in non-ICT assets and labor quality contributed 37 percent and 10 percent, respectively, to the declining trend in productivity growth between the two periods.

©International Monetary Fund. Not for Redistribution 62 German Productivity Growth: An Industry Perspective

TABLE 3.2 Contributions to Growth of GDP (Annual average growth rates, in percentage pointsa) Germany United States 1990–1995 1995–2004 2004–2007 1990–1995 1995–2004 2004–2007 1GDP (2)+(3)+(4)+(7) 2.2 1.4 2.1 2.5 3.3 2.5 Contributions from 2 Hours worked –0.4 –0.2 0.3 0.8 0.7 1.7 3 Labor quality 0.2 0.1 –0.1 0.2 0.2 0.3 4 Capital stock (6)+(7) 1.0 0.7 0.8 1.0 1.5 0.9 5 ICT capital 0.3 0.4 0.3 0.6 1.0 0.4 6 Non-ICT capital 0.7 0.3 0.5 0.4 0.6 0.5 7 TFP 1.4 0.7 1.2 0.5 0.9 –0.5

Memo item: Labor productivity 2.6 1.6 1.8 1.7 2.7 0.8 growth

Source: The Conference Board Total Economy Database, September 2010; and IMF staff estimates. a Based on the difference in the log of the levels of each variable.

Jorgenson, Ho, and Stiroh (2004), Triplett and Bosworth (2006), Bosworth and Triplett (2007), and more recently Jorgenson, Ho, and Stiroh (2008). In 2005, a productivity catch-up resumed, mainly reflecting a U.S. slowdown. Germany’s annual labor productivity growth accelerated only moderately (by 0.2 percent) in 2004–2007 relative to 1995–2004, as a pick-up in employment offset higher TFP growth (by 0.5 percent) and a higher contribution of non-ICT capital investment (by 0.2 percent). However, since TFP is measured as a residual, the recent turnaround could be largely due to unmeasured cyclical factors—including higher capacity utilization immediately precrisis in Germany. Moreover, unlike in the United States in the mid-1990s, the recent productivity acceleration in Germany was not accompanied by higher investment in ICT assets, which also suggests that cyclical factors might have been at play rather than a structural shift to a more “knowledge-driven” economy. The results discussed so far apply to the total economy, including both public and private services. When public services—including health, education, and other public services—are excluded, the productivity growth gaps between Germany and the United States that emerged in the mid-1990s are magnified and TFP, rather than capital accumulation, explains the bulk of the gap.9 Germany’s labor productivity in the private economy grew at broadly the same rate during 1995–2004 as the total economy (including public services) grew, namely at 1.6 percent (Table 3.3). The acceleration in U.S. productivity growth therefore becomes even more striking when focusing only on the private economy: productivity

9 Other excluded services are public administration and defense. Following van Ark, O’Mahony, and Timmer (2008), we also exclude real estate (ISIC 70), because output in this industry mostly reflects imputed housing rents rather than sales of firms.

©International Monetary Fund. Not for Redistribution Poirson 63 growth in the private economy there rose by 3.1 percent annually over that period, compared to 2.7 percent for the total economy. The U.S. productivity lead is thus especially pronounced in the private economy. Specifically, we find that: • The annual productivity gap with the United States during 1995–2004 for only the private economy is 1.5 percent, compared to 1.1 percent for the total economy (including public services). • Since 2005, the results continue to show the resumption of a productivity catch-up when only the private economy is considered. During 2004– 2007, the German resurgence is more pronounced when public services are excluded, and the U.S. slowdown is less extreme. Both trends suggest that a catch-up process has resumed more recently, although as noted above it is too early to conclude that the overall lag in German productivity growth has reversed, given the importance of cyclical factors over this relatively short period and the lack of a pick-up in ICT investments. • TFP growth drives the differences between the two countries’ private- economy productivity growth rates during 1995–2004, contributing over two-thirds (72 percent) of the productivity gap (compared to 15 percent for the total economy, including public services). The effect of slower investment in ICT capital on Germany’s private economy is also an important factor, accounting for more than a third (35 percent) of the slower labor productivity growth in Germany relative to the United States over that period (compared to more than half when the total economy is considered). The contribution of differences in the pace of labor skill growth increases marginally (to 12 percent of total) when only the private economy is considered.10 • When focusing on developments over time in the two countries, a U.S. slowdown, rather than significantly higher productivity growth in Germany, continues to account for most of the recent turnaround since 2005. Within Germany’s private economy, the pick-up in TFP growth (by 1.2 percent) since 2005 is more pronounced than for the total economy including public services (by 0.5 percent). Similar to the result for the total economy, the contribution of ICT capital accumulation to productivity growth in the private economy fails to accelerate in 2004–2007, unlike the U.S. experience of the mid-1990s, suggesting that recent developments may reflect

10 The overall contribution of human capital would be underestimated in a growth-accounting framework if human capital mainly influences labor productivity growth through its impact on TFP growth. Higher education, in particular, arguably has a supportive role in fostering technological improvements and enabling a fast adjustment to new technologies. Using a sample of 13 EU countries plus the U.S., Molagoda and Perez (2011) confirm empirically the significant role of human capital in explaining TFP differences across countries through both its direct impact on innovation by the productivity leader and indirectly by increasing the size of knowledge spillovers. The largest impact of human capital is found for countries at or close to the productivity frontier, possibly because innovation is a relatively more skill-intensive activity than imitation. See also Vandenbussche, Aghion, and Meghir (2006).

©International Monetary Fund. Not for Redistribution 64 German Productivity Growth: An Industry Perspective

TABLE 3.3 Contributions to Growth of Real Output in the Market Economy (Annual average growth rates, percentage points) Germany United States 1980–1995 1995–2004 2004–2007 1980–1995 1995–2004 2004–2007 1GDP (2)+(3) 1.9 1.1 2.4 3.3 3.8 2.8 2 Hours worked –0.5 –0.6 0.5 1.3 0.7 1.4 3 Labor productivity 2.4 1.6 1.9 2.0 3.1 1.4 (4)+(5)+(8) Contributions from 4 Labor composition 0.2 0.1 –0.2 0.2 0.3 0.1 5 Capital stock per 1.3 1.1 0.6 1.0 1.4 0.6 hour (6)+(7) 6 ICT capital per 0.3 0.5 0.4 0.7 1.0 0.5 hour 7 Non-ICT capital per 1.0 0.6 0.2 0.3 0.4 0.1 hour 8 TFP 0.8 0.4 1.6 0.8 1.4 0.6

Labor productivity 1.4 1.1 1.7 1.7 2.8 1.2 contribution from the knowledge economy (4)+(6)+(8)

Sources: EU Klems database, November 2009 release; and IMF staff estimates. Note: ICT: information and communications technology; TFP: total factor productivity. Numbers may not sum exactly due to rounding.

temporary cyclical factors rather than the emergence of a knowledge economy typically associated with both higher investment in ICT assets and higher TFP growth. The results above are broadly consistent with existing empirical evidence suggesting that slow increases in ICT investments and TFP growth were the two most important contributors to the aggregate productivity gap with the United States in the mid-1990s. According to Molagoda and Perez (2011), these two factors account for 42 percent and 33 percent, respectively, of the German-U.S. labor productivity differential in the private economy over the 1995–2007 precrisis period.11 Based on their results, similar findings would be obtained using the United Kingdom or Sweden as productivity leader benchmarks instead of the United States.

AN INDUSTRY PERSPECTIVE This section shifts from the aggregate perspective of the previous section to a sector-level (more disaggregated) perspective, allowing us to document the contributions of each sector to aggregate labor productivity growth and the key sources of productivity growth at the sector and industry level. Based on industry-level

11 Molagoda and Perez (2011) do not provide the breakdown over 1995–2004 and 2004–07, only the full sample average results.

©International Monetary Fund. Not for Redistribution Poirson 65

200 Germany United States 180

160

140

120

100

40 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Figure 3.3 Value Added in the Services Sector (Gross value added, volume indices, 1995=100) Sources: EU Klems database, November 2009 release; IMF staff estimates.

800 Germany United States 700

600

500

400

300

200

100

0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Figure 3.4 ICT Capital Services (Volume indices, 1995=100) Sources: EU KLEMS database, November 2009 release; IMF staff estimates. measures of output, inputs, and TFP from the EU KLEMS database, the results show that productivity differences are strong across industries too. In the case of the German-U.S. differential in productivity growth, we find that private services and, to a lesser extent, ICT production are the main sectors accounting for the aggregate productivity gaps in the private economy (Figures 3.3 and 3.4).12 Private services (including wholesale and retail trade, hotels and restaurants, transport services, and financial and business services) also explain the bulk of

12 ICT production includes production of electrical machinery and telecommunication services.

©International Monetary Fund. Not for Redistribution 66 German Productivity Growth: An Industry Perspective

BOX 3.3 The Rising Importance of Service Sectors

Both Germany and the United States have experienced a major shift of production and employment from manufacturing and other goods-producing industries (such as agriculture and mining) toward services. Over the period 1980–2007, the share of hours worked in manufacturing declined by more than 30 percent in Germany and was almost halved in the United States. Private services—including trade and transportation services, hotels and restaurants, and business and financial services—now account for 60 percent of private employment in Germany (compared to over two-thirds in the United States). In the United States, the number of hours worked is now nearly five times larger than in market services than in manufacturing. Even in Germany, where manufacturing plays an important role (accounting for 28 percent of total private output), the number of hours worked in market services is almost three times greater than that for manufacturing. This shift has important potential implications for productivity growth. Historically, innovation and technical change in manufacturing have been the central source of productivity growth. With the emergence of the knowledge economy in the mid-1990s, the center of gravity of productivity improvements may have shifted to service industries. The experience of the United States in the mid-1990s bears out this hypothesis. The U.S. revival reflected the rising growth contributions of ICT productivity and investment, especially in the distribution (trade and transportation) and other private services sectors. Due to the rising importance of the service sectors, durable productivity gains in Germany will need to extend to these industries in order to have a lasting impact on both real incomes and potential growth.

Germany’s long-term declining trend in productivity growth since the mid- 1990s, and the reverse trend in the United States over the same period (Box 3.3, Table 3.4, and Figure 3.5). We also find that productivity gaps in the private services industries largely reflect slower investment in ICT assets and the atten- dant slower TFP growth. By contrast, productivity in the goods production sectors (including agriculture, mining, and manufacturing other than electrical machinery, utilities, and construction) has consistently been higher in Germany than in the United States. This principally reflects the traditional German strength in manufacturing, a

TABLE 3.4 Major Sector Contribution to Average Annual Labor Productivity Growth in the Market Economy (Annual average growth rates, percentage points) Germany United States 1980–1995 1995–2004 2004–2007 1980–1995 1995–2004 2004–2007 1 Market economy 2.4 1.6 1.9 2.0 3.1 1.4 (2)+(3)+(4)+(5) 2 ICT production 0.3 0.5 0.5 0.4 0.9 0.7 3 Goods production 1.1 1.1 0.6 0.8 0.5 –0.1 4 Market services 0.9 0.2 0.9 0.8 1.9 0.9 5 Reallocation 0.0 –0.1 –0.2 –0.1 –0.2 –0.2

Source: EU KLEMS database, November 2009 release and IMF staff estimates. Notes: ICT: information and communications technology. “Reallocation” refers to labor productivity effects of reallocation of labor between sectors. ICT production includes manufacturing of electrical machinery and post and telecommuni- cations services. Goods production includes agriculture, mining, manufacturing (excluding electrical machinery), construction, and utilities. Market services include distribution services; financial and business services, excluding real estate; and personal services. Numbers may not sum exactly due to rounding.

©International Monetary Fund. Not for Redistribution Poirson 67

Germany's private sector fell behind in the mid-nineties. Although productivity in the goods sector remained strong. . . 4.5 Private Economy: Productivity Per Hour Worked 4.0 Goods Production: Productivity Per Hour Worked (Five-Year Average Annual Percent Change) 4.0 3.5 (Five-Year Average Annual Percent Change) 3.5 3.0 3.0 2.5 2.5 2.0 2.0 1.5 1.5 1.0 1.0 Germany Germany 0.5 United States 0.5 United States 0.0 0.0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

. . .it lagged in information and communication technologies (ICT)...... and it was especially slow in the services sector.

14 ICT Production: Productivity Per Hour Worked 4.0 Private Services: Productivity Per Hour Worked (Five-Year Average Annual Percent Change) (Five-Year Average Annual Percent Change) 12 3.5 3.0 10 2.5 8 2.0 6 1.5

4 1.0

2 Germany 0.5 Germany United States United States 0 0.0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Total factor productivity (TFP) led the US revival since . . .while German TFP lagged until just before the crisis. 1995. . . United States: Private Economy Productivity Growth Germany: Private Economy Productivity Growth 3.5 Decomposition (Five-Year Average Contribution, 3.5 Decomposition (Five-Year Average Contribution, in Percent) in Percent) 3.0 3.0

2.5 2.5

2.0 2.0

1.5 Factor Intensity 1.5 TFP Factor Intensity 1.0 1.0

0.5 0.5 TFP

0.0 0.0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Figure 3.5 Germany: Productivity Trends in the Private Economy, 1985–2007 Sources: EU KLEMS database; and IMF staff estimates. strength that makes Germany a formidable exporter of a wide range of precision manufactured products. Productivity growth in the ICT production sectors (including production of electrical machinery and telecommunication services) has been rising in line with worldwide trends but has lagged the United States since the mid-1980s (Figure 3.5). Within the services sector, German-U.S. productivity differentials are concentrated in a handful of industries. Productivity gaps during 1995–2004 were especially large in distribution services (including trade and transportation) and in financial and business services. The distribution sector contributed a third of the overall annual productivity gap of 3 percent during 1995–2004. In finance and business services, the gap was even bigger, at almost 2 percent (nearly two- thirds of the total). In both cases, productivity growth differences reflect to a large

©International Monetary Fund. Not for Redistribution 68 German Productivity Growth: An Industry Perspective

TABLE 3.5 Contributions of Sectors to Average Annual Labor Productivity Growth in Market Services (Percentage points) Germany United States 1980–1995 1995–2004 2004–2007 1980–1995 1995–2004 2004–2007 Market services labor 2.1 0.3 1.7 1.6 3.3 1.6 productivity

Distribution services 1.0 0.9 1.0 1.1 1.9 0.7 contribution from factor intensity 0.4 0.3 0.2 0.4 0.6 0.3 growth from TFP growth 0.6 0.6 0.7 0.7 1.3 0.4

Finance and Business 0.6 –0.7 0.4 0.0 1.2 0.2 services contribution from factor intensity 1.4 1.4 0.7 0.9 1.2 0.3 growth from TFP growth –0.8 –2.2 –0.2 –1.0 0.0 –0.1

Personal services 0.1 –0.2 0.1 0.2 0.1 0.3 contribution from factor intensity 0.2 0.0 0.0 0.0 0.1 0.1 growth from TFP growth –0.2 –0.2 0.1 0.1 0.1 0.3

Contribution from 0.4 0.3 0.2 0.3 0.0 0.3 labor reallocation

Sources: EU KLEMS database, November 2009 release; and IMF staff estimates. Note: Factor intensity relates to the total contribution from changes in labor composition and in capital deepening of information and communications technology (ICT) and non-ICT assets. The reallocation effect refers to the impact of changes in the distribution of labor input between industries on labor productivity growth in market services. Numbers may not add up due to rounding. TFP: total factor productivity.

extent differences in TFP growth at the sectoral level (i.e., efficiency of input use) rather than differences in factor intensity growth (i.e., growth in both human and physical capital inputs). The closure of the gap vis-à-vis the United States between 2005 and 2007 mainly reflected slower productivity growth in the United States’ distribution and financial and business services and improved performance in Germany’s financial and business services. In Germany, the turnaround in financial and business services reflects higher TFP growth rather than larger contributions of capital accumulation and labor quality. However, international experience suggests that sustaining this trend will require concurrent improvements in TFP and investment—especially in ICT assets—since both are drivers of innovation effects on services productivity. The continuing productivity gaps in the services sector reflect in part Germany’s lag, by advanced country standards, in the use of ICT. Compared to some other advanced economies, Germany lags in research and use of ICT. This is evidenced by the low share of ICT in total research and development spending, compared to the average in OECD countries, and by the relatively low extent of

©International Monetary Fund. Not for Redistribution Poirson 69

70 Germany United States OECD total 60

0 2000 2001 2002 2003 2004 2005 2006 2007

Figure 3.6 Internet Subscribers (Percent of population) Source: Organisation for Economic Co-operation and Development. business and government internet use. The latter in turn seems caused by several infrastructure and financing factors, including: • a relatively low number of secure internet servers; • low public investment in advanced technologies compared to some other OECD countries; and • a comparatively low willingness of the financial system to provide financing for high-risk, innovative projects (Box 3.4). Internet penetration in Germany (measured by the number of subscribers as a share of total population) is in line with the average for OECD countries, but it is lower than in the United States (Figure 3.6). A limited initial public offering (IPO) market and tax obstacles appear to be contributing to the limited development of German private equity and venture capital markets and thus to the limited availability of risk financing. Difficulty in achieving successful exits, lack of entrepreneurial talent, and unfavorable tax policies (including tax obstacles to cross-border investments and double taxation) are the main unfavorable climate factors for venture capital in Germany, according to a 2010 survey of 516 firms in nine countries: 72 percent of respondents cited lack of entrepreneurial talent and tax policies as the main obstacles, and 67 percent considered exit difficulties the main issue. Eliminating tax obstacles to cross-border investments and double taxation would require coordinated action at the EU level to remove these and other regulatory barriers (e.g., separate regis- tration requirements), thus allowing even smaller funds to invest EU-wide more efficiently, develop specialized sectoral expertise, and reap economies of scale.13

13 While there is a consensus among the member states on promoting mutual recognition of national frameworks, no significant measures have yet been taken that would make fundraising and investing across borders easier (European Commission Enterprise and Industry Directorate General, 2009).

©International Monetary Fund. Not for Redistribution 70 German Productivity Growth: An Industry Perspective

BOX 3.4 Productivity Growth in the Information Age

International experience suggests that innovation and productivity growth in services are closely related to the wider use of information and communication technology (ICT). In the U.S. private services sector since 1995, high levels of investment in ICT have been followed by rapid productivity growth. Other high-performing economies, such as Sweden, Finland, and the U.K. also have relatively high ICT uptake; in addition, the high- productivity countries are characterized by a relatively high level of human capital employed in the services sector (Molagoda and Perez, 2011). Germany lags some advanced countries both in the use of ICT and in human capital in the services sector. Government usage of ICT and the number of secure servers are in the mid- to lower-range. The ratio of high-skilled labor employed in the services industries in 2007 is below 20 percent, compared to almost 50 percent for the U.S. and more than 40 percent for Finland (Molagoda and Perez, 2011). In the group of advanced nations, Germany’s ICT research and development share in total research and development also falls in the mid- to lower-range. While public policy to support ICT development and use remains controversial, the evidence suggests that ICT development and use can be fos- tered through innovation policies, including promoting the availability of risk capital and increasing the commercial use of intellectual property rights held by universities and research institutions. A common European market for services would also generate economies of scale and raise incentives to invest in new technologies.

The share of investment and research in information The number of secure servers is low relative technologies is relatively small in Germany. to peers. 70 6.0 ICT R&D expenditure as a share of total R&D, 2007 KOR USA 60 (Percent) 5.5 SWECAN CHE DEU NLD 50 FIN UK DNK 5.0 FRA JPN 40 AUS BEL AUT NOR 4.5 30 ESP 20 4.0 PRT ITA 10

ICT Usage (index 1–7) 3.5 0 GRC 3.0 Korea OECD Finland Sweden 0 500 1000 1500 Germany Netherlands Secure internet servers per milllion population, 2008 United States

Public procurement of high-tech products is High-risk financing is less available than in low by advanced-economy standards. other OECD countries. 6.0 6.0 KOR KOR CHE CHE USA 5.5 SWEUSA 5.5 SWE UK UK CAN NLD CAN DNK DEU NLD DEU DNK FIN 5.0 JPN FRA 5.0 JPN FRA NOR FIN AUT AUS NOR AUT AUS BEL BEL 4.5 4.5 ESP ESP 4.0 PRT 4.0 PRT ITA

ICT Usage (index 1–7) 3.5 ICT Usage (index 1–7) 3.5 GRC GRC 3.0 3.0 2.5 3.0 3.5 4.0 4.5 5.0 2.5 3.0 3.5 4.0 4.5 5.0 Government procurement of advanced Venture Capital Availability (survey index 1–7) technology products (survey index 1–7)

Figure 3.4.1 Research and Use of Information and Communication Technology Sources: The Global Information Technology Report, 2009–10; and Organisation for Economic Cooperation and Development.

©International Monetary Fund. Not for Redistribution Poirson 71

The lack of exit opportunities in Germany reflects the small share of global IPO listings in Deutsche Boerse.14 Reasons for a limited IPO market in Germany compared to other key financial centers seem related to market liquidity: a shal- lower pool of equity capital compared to the United States, the United Kingdom, or the countries associated with the Euronext exchange, and a narrower investor base compared to the United States or the UK. It is these factors that limit the IPO market, rather than cost factors such as listing requirements and underwrit- ing fees, where Germany appears very competitive (Oxera, 2006). Another reason for Germany’s delay in adopting the Internet economy and constraining factors for innovation is deficiency in the insolvency law, which results in relatively less creditor-friendly legislation than in the United Kingdom and thus less willingness of banks to lend to high-risk projects. Specifically, German banks face costlier and lengthier proceedings relative to the United Kingdom and thus potentially higher legislation-induced credit risk (Schmieder and Schmieder, 2011). In response, they demand relatively more credit risk miti- gation than U.K. and U.S. banks do, but still recover less than do U.K. banks. Banks’ lower willingness to lend to risky projects in turn could hinder entrepreneurship and risk-taking. To be on par with U.K. banks, Schmieder and Schmieder (2011) calculate that formal bankruptcy proceedings in Germany would have to be shortened by about one half. Relatively limited use of public procurement to provide sustained ICT demand in Germany, as compared to other advanced countries such as Korea, Sweden, and the United States, may be another factor explaining the comparatively low investment in new, high-risk technologies. The premise is that there are positive externalities in the use of ICT caused, for example, by network effects or complementary investments, such as organizational change, that go unmeasured, and large fixed costs from the required infrastructure investments (servers, infrastructure software, storage).15 Business models for ICT companies or services companies that make heavy use of ICT are thus reliant on scale to be profitable. A transparent procurement process that enables greater usage based on open standards can create synergies. Greater European integration in services provision would also help raise investment in new, high-risk technologies, It would generate economies of scale and raise incentives to invest in and provide risk financing for innovative applications of ICT in services. Obstacles to an integrated European market for online and off-line services include a variety of compliance and regulatory issues,

14 The number of IPOs expected in Germany in 2011 is 20, compared to a backlog of 150 deals in the U.S at end-February; Deutsche Boerse lagged other IPO markets in China, London, New York, Tokyo, Mumbai, Australia, and Korea in 2010, both in number of deals and funds raised (Ernst & Young, 2011). 15 See, for example, Stiroh (2002) and McKinsey Global Institute (2010). While the potential offered and challenges posed in using public procurement as an instrument for innovation have been largely ignored or downplayed, a number of empirical studies conclude that over long time periods, state procurement triggered greater innovation impulses in more areas than did research and development subsidies (see Edler and Georghiou, 2007, for a review of the evidence).

©International Monetary Fund. Not for Redistribution 72 German Productivity Growth: An Industry Perspective

including country-specific consumer-protection laws, VAT rules, electronic waste regulations, varying copyright rules, and postal systems. Moreover, the need to tailor products to each national market requires maintaining region- specific teams and translating every item, website, and online application, in several different languages. Finally, price-comparison websites, which make online shopping attractive in the United States, are relatively few in Europe, where existing sites remain national.16 While these obstacles can be overcome,17 they result in higher costs of doing business for services firms that attempt to operate across borders, and they further add to barriers to entry for smaller, start- up firms that do not have the administrative and personnel resources to cope with different national regulations. Closing the productivity gap with the United States in private services could raise TFP growth, and thus potential growth, by up to 0.6 percentage points annually. The overall impact on potential growth could be higher if the TFP increase is associated with higher investment in new technologies, as was the case in the United States in the mid-1990s. Shifting adoption patterns for new IT services and products, such as cloud computing (forecast to grow at more than five times the rate of traditional IT products), suggest that the share of these growing markets, currently heavily concentrated in the United States at 70.2 percent, is likely to decline in the United States to 51.4 percent by 2014, while the share of Western Europe and Asia/Pacific, excluding Japan, is set to grow rapidly (IDC, 2010). Countries that plan for this shift by creating a favorable market environment, in terms of availability of risk financing and significant additional investment in broadband and complementary infrastructure, are likely to benefit disproportionately from this catch-up effect.18 EU regulations that hinder the development of new IT service industries, such as EU laws that restrict the movement of data and access to databases across borders, may also need to be revisited, for example to allow storage of data by companies based in the EU with vendors based outside the region.

CONCLUSION International experience suggests that ICT applications in traditionally low-tech industries such as the retail and wholesale sectors lead to technology and innovation effects, which in turn raise TFP growth. While large and internationally

16 See The Economist, October 28, 2010. 17 Successful internet start-ups, such as Berlin-based Wooga, illustrate the scope for exploiting the synergies of a European user market. 18 Much of the available empirical research refers to the impact of the Internet or the “digital economy” rather than ICT infrastructure. A recent study using data for the 48 U.S. states over 2003–05 suggests that for every 1 percentage point increase in broadband penetration in a particular area, employment growth is estimated to increase 0.2–0.3 percentage points per year (Crandall, Litan, and Lehr, 2007). The study also finds that state output is positively associated with broadband use, although the impact is not statistically significant.

©International Monetary Fund. Not for Redistribution Poirson 73 connected German companies are perceived to be at the forefront of innovation and usage of new technologies, overall ICT uptake and the prioritization of ICT in national growth strategy are both low in Germany relative to some other advanced countries. In this paper we argue that policy measures are needed to raise the low annual TFP growth rate of only 0.5 percent during the last decade (see Chapter 2). While the more recent pick-up in productivity growth in Germany since 2005 is encouraging, the improvement could be largely cyclical, since it has not been accompanied by a concomitant rise in ICT investments. The experience of other fast-growing advanced countries suggests that to ensure a sustained rise in productivity and TFP growth in the private economy and to close the productivity gaps in the service sectors, Germany needs to encourage more widespread usage of ICT and human capital as well as investment and innovation, particularly in the private services sector (i.e., outside Germany’s traditional strengths). The required measures to raise incentives to invest in higher-risk, higher-growth sectors include: 1. further development of venture capital and private equity markets (backed by a more efficient insolvency process); 2. increased commercial use of intellectual property rights held by universities and research institutions; 3. removal of uncertainties regarding tax treatment; and 4. further European integration in services provision. Strong supply-side policies alone—including improvements of insolvency law system and the financing environment—are important to kick-start innovation and encourage risk-taking, but they may not be sufficient to sustainably raise incentives to invest in new technologies. Some have argued that ICT uptake is also importantly shaped by the availability of a large enough market due to the presence of network effects and large fixed costs to IT infrastructure investments. In Germany’s case, there may be scope for greater but transparent public procurement of high-tech products to support overall incentives for innovation and the development of new ICT applications. Greater harmonization of rules and regulations to foster greater European integration in services provision could help raise competition among service providers and thus lower prices and increase incentives for ICT investments and ICT usage, in a more efficient way than targeted subsidies. A common European market for services would also enable the realization of economies of scale, with expected positive effects on TFP growth and investment incentives.

REFERENCES Allard, Céline, and Luc Everaert, 2010, “Lifing Euro Area Growth: Priorities for Structural Reforms and Governance,” IMF Staff Position Note SPN/10/19 (Washington, DC: International Monetary Fund).

©International Monetary Fund. Not for Redistribution 74 German Productivity Growth: An Industry Perspective

Bartelsman, Eric, Andrea Bassanini, John Haltiwanger, Ron Jarmin, Stefano Scarpetta, and Thorsten Schank, 2008, “The Spread of ICT and Productivity Growth: Is Europe Really Lagging Behind in the New Economy?,” unpublished paper, OECD. Bosworth, Barry P., and Jack E. Triplett, 2007, “The Early 21st Century Productivity Expansion is Still in Services,” International Productivity Monitor, 14 (Spring). The Conference Board, 2009, “Productivity, Performance, and Progress: Germany in International Comparative Perspective.” Available online at http://www.conference-board. org/publications/publicationdetail.cfm?publicationid=1682. Crandall, Robert W., Robert E. Litan, and William Lehr, 2007, “The Effects of Broadband Deployment on Output and Employment: a Cross-Sectional Analysis of U.S. Data,” Issues in Economic Policy, No. 6 (Washington, DC: The Brookings Institution). Deloitte, 2010, Results from the 2010 Global Venture Capital Survey, available at www.nvca.org. The Economist, 2010, “Europe’s Need for E-Freedom,” available at http://www.economist .com/node/17361454. Edler, Jakob, and Luke Georghiou, 2007, “Public Procurement and Innovation—Resurrecting the Demand Side,” Research Policy, Vol. 36, 949–63. Ernst & Young, 2011, Global IPO Trends, available at http://www.ey.com/GL/en/Services/ Strategic-Growth-Markets/Global-IPO-trends-2011. European Commission Enterprise and Industry Directorate General, 2009, Cross-Border Venture Capital in the European Union: European Commission Work on Removing Obstacles (Brussels: European Commission). EU ICT Task Force, 2006, “Fostering the Competitiveness of Europe’s ICT Industry.” Available online at http://ec.europa.eu/information_society/eeurope/i2010/docs/high_level_group/ ict_task_force_report_nov2006.pdf. IDC, 2010, Worldwide and Regional Public IT Cloud Services 2010–14 Forecast. Inklaar, Robert, Marcel P. Timmer, and Bart van Ark, 2008, “Market Services Productivity,” Economic Policy, January, pp. 139–94. Jorgenson, Dale W., Mun S. Ho, and Kevin J. Stiroh, 2004, “Will the U.S. Productivity Resurgence Continue,” Current Issues in Economics and Finance, Vol. 10, No. 13 (New York: Federal Reserve Bank of New York). ———, 2008, “A Retrospective Look at the U.S. Productivity Growth Resurgence,” Journal of Economic Perspectives, Vol. 22, No. 1, pp. 3–24. McKinsey Global Institute, 2010, “Beyond Austerity: A Path to Economic Growth and Renewal in Europe” (McKinsey & Company). ———, 2011, “European Growth and Renewal: The Path from Crisis to Recovery, Updated Research” (McKinsey & Company). Molagoda, Nandaka, and Esther Perez, 2011, “Raising Potential Growth in Europe: Mind the Residual,” Chapter V in Euro Area Policies—Selected Issues, SM/11/160, pp. 62–79 (Washington: International Monetary Fund). Nahuys, Richard, and Henry van der Wiel, 2005, “How Should Europe’s ICT Ambitions Look Like? An Interpretative Review of the Facts,” Discussion Paper No. 05–22 (Utrecht: Tjalling C. Koopmans Research Institute.) Oxera Consulting Ltd, 2006, The Cost of Capital: An International Comparison). Available at http://www.nd.edu/~carecob/May2008Conference/Papers/OxeraCostofcapitalreport ExecSummary.pdf. Schmieder, Christian, and Philipp Schmieder, 2011, “The Impact of Legislation on Credit Risk—Comparative Evidence from the United States, the United Kingdom, and Germany,” IMF Working Paper No. WP/11/55 (Washington, DC: International Monetary Fund). Stiroh, Kevin J., 2002, “Information Technology and the U.S. Productivity Revival: What Do the Industry Data Say?” American Economic Review, Vol. 92, No. 5, pp. 1559–76. Timmer, Marcel P., Gerard Ypma, and Bart van der Ark, 2003, “IT in the European Union: driving productivity divergence?” GGDC Research Memorandum 200363, Groningen Growth and Development Centre, University of Groningen.

©International Monetary Fund. Not for Redistribution Poirson 75

Triplett, Jack E., and Barry P. Bosworth, 2006, “Baumol’s Disease Has Been Cured: IT and Multifactor Productivity in U.S. Services Industries,” in The New Economy and Beyond: Past, Present, and Future, ed. by Dennis W. Jansen (Cheltenham, U.K.: Edgar Elgar). van Ark, Bart, 2010, “Productivity, Sources of Growth, Potential Output in the Euro Area and the United States,” Intereconomics, Vol. 1, pp. 17–20. van Ark, Bart, Mary O’Mahony, and Marcel P. Timmer, 2008, “The Productivity Gap Between Europe and the United States: Trends and Causes,” Journal of Economic Perspectives, Vol. 22, No. 1, pp. 25–44. Vandenbussche, Jérôme, Philippe Aghion, and Costas Meghir, 2006, “Growth, Distance to Frontier, and Composition of Human Capital,” Journal of Economic Growth, Vol. 11, No. 2, pp. 97–127. Voss, Romy, and Christoph Müller, 2009, “How Are the Conditions for High-Tech Start-Ups in Germany?,” International Journal of Entrepreneurship and Small Business, Vol. 7, No. 3, pp. 284–311.

©International Monetary Fund. Not for Redistribution 76 German Productivity Growth: An Industry Perspective

APPENDIX

TABLE 3A.1 Germany and the United States: Average Annual Growth Rates of GDP, G DP per Capita, and GDP per Hour Worked, 1950–2009 (Percent) Average annual growth in GDP GDP per capita GDP per hour worked 1950–1973 Germany 6.0 5.3 5.8 US 3.9 2.5 2.6 1973–1995 Germany 2.1 1.8 2.8 US 2.9 1.7 1.2 1980–1995 Germany 2.1 1.8 2.8 US 3.0 1.9 1.4 1995–2004 Germany 1.4 1.3 1.7 US 3.4 2.3 2.4 2004–2007 Germany 2.1 2.1 1.7 US 2.6 1.6 1.0 2008 Germany 1.3 1.3 0.1 US 0.0 –0.9 0.8 2009 Germany –4.9 –4.9 –2.4 US –2.6 –3.5 2.5

Sources: The Conference Board Total Economy Database, September 2010; and IMF staff estimates. Note: Germany’s pre-1988 population is estimated as the sum of West Germany and East Germany. Relative levels are based on Elteto-Koves-Szulc (EKS) purchasing power parities for GDP for 2009.

©International Monetary Fund. Not for Redistribution CHAPTER 4 What Does the Crisis Tell Us about the German Labor Market?

MARTIN SCHINDLER

Germany’s employment fell only marginally during the crisis, despite a sharp drop in output and in contrast to historical employment patterns over the business cycle. This paper offers new perspectives on the apparent “puzzle” of the German labor market response. Labor market developments during the crisis are consistent with a view that Germany experienced mainly a temporary demand shock, necessitating relatively little sectoral reallocation and with employers therefore adjusting hours of work while retaining workers. The labor market dynamics also provide evidence that past reforms, including especially the Hartz reforms during the early 2000s, enabled the labor market to react in a way that previously had not been possible to the same extent. These reforms facilitated a shift from adjustment in the number of individuals employed to adjustments in hours. Thus, while the employment decline was smaller than expected based on past dynamics, the reduction in hours worked was more in line with the drop in output. And lastly, the moderate employment decline during the crisis reflected the longer-term structural changes in the labor market. The upward momentum in employment stemming from the earlier reforms held back the adverse crisis impact.

INTRODUCTION The German labor market has gone through a remarkable transformation over the past two decades. From being labeled the “sick man of Europe” as unemployment rates steadily went up during the 1990s, and even reached double-digit territory during the mid-2000s, to references to the “German labor market miracle” during the recent financial crisis when unemployment, after a brief rise at its onset, continued the downward trend that had started around 2005. This transformation is particularly striking in comparative perspective: Germany was one of the European countries that motivated Ljungqvist and Sargent’s (1998) study of how and why the U.S. labor market performed so much better, especially in turbulent times. Now, economists are puzzling over why the German

The author has benefitted from comments and suggestions by Ashoka Mody, Helène Poirson, Fabian Bornhorst and participants at the Germany in an Interconnected World conference at the Ministry of Finance in Berlin (May 2011), especially the paper’s discussant, Werner Eichhorst. Susan Becker provided excellent research assistance. 77

©International Monetary Fund. Not for Redistribution 78 What Does the Crisis Tell Us about the German Labor Market?

labor market has proven to be so much more resilient than the U.S. market, even worrying about a jobless recovery there while German unemployment continues to decline. This chapter aims to provide new perspectives on the recent labor market dynamics and to understand how “miraculous” the crisis performance actually was. It also aims to answer the question of what the crisis can tell us about the German labor market, and in particular whether the recent experience reflects something more fundamental. The analysis suggests that, broadly speaking, the labor market response has been benign, but less so than meets the eye, and that the German labor market does indeed appear to be structurally different from what it was even a decade ago. The chapter proceeds by providing some general background on longer-term and more recent developments, discussing a number of factors that help to better understand the recent labor dynamics. The appendix provides details on the labor market simulations reported here.

BACKGROUND Germany has long been labeled the “sick man” of Europe, owing in large part to its ailing and sluggish labor market with its high and increasing unemployment (e.g., The Economist, 1999). Indeed, starting from an unemployment rate of less than one percent in 1970, German unemployment has almost continuously trended upwards, interrupted only by brief cyclical declines in unemployment, such as those during 1976–1980, 1985–1990, 1997–2001, and the most recent decline since 2005 (with only a moderate uptick in 2009). In each of those cases, the decline in unemployment was paralleled by a growth recovery in output (Figure 4.1). 12

4 West Germany Unified Germany

0 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 Figure 4.1 Germany: Unemployment Rate (National definition and International Labour Organization, percent) Source: German Statistical Office.

©International Monetary Fund. Not for Redistribution Schindler 79

France Italy Japan United Kingdom 15 United States Germany 12

−3

−6 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 Figure 4.2 Unemployment in Advanced Economies (ILO definition, percent) Source: International Labour Organization (ILO).

However, the cyclical upswings in GDP were not sufficient to turn around the secular increase in unemployment. With the exception of the most recent (ongoing) episode, each trough in unemployment was higher than the previous one. Thus, from nearly zero unemployment in the early 1970s, German unemployment peaked in 2005 at over 11 percent. Notably, the deterioration in labor market outcomes was somewhat unique to Germany (Figure 4.2). Along with Japan, Germany had one of the lowest unemployment rates among the large advanced economies in the 1970s, substantially lower than in the United States, where unemployment was on the order of 6 percent at that time. However, while U.S. unemployment peaked at nearly 10 percent in 1982 and declined thereafter, German unemployment continued to rise. By the early 2000s, Germany had one of the highest unemployment rates among its peers. The long upward trend that began as early as the 1970s suggests that Germany’s labor market problems were not specific to the structural challenges posted by German unification in 1990. A broad consensus has emerged on the main underlying causes. Siebert (2003) views the secular step-wise increase in unemployment as the result of weak job creation, in turn caused by the German institutional design for wage formation, which at times has led to binding negotiated wage increases and insufficient wage differentiation; and as the result of a high reservation wage, resulting from high unemployment and social welfare benefits.1

1 This notion of “binding wage increases” is based on the observed decline in the “wage drift” (the difference between actual and negotiated wage increases), especially during the 1990s (Siebert, 2003, p. 8).

©International Monetary Fund. Not for Redistribution 80 What Does the Crisis Tell Us about the German Labor Market?

With its combination of high unemployment and welfare benefits, rigid wage determination mechanisms, and high firing costs, Germany was typical of what many authors saw as the main differences between European and U.S. labor market models (Siebert, 1997; Ljungqvist and Sargent, 1998). In addition, Prescott (2004) and Rogerson (2010) cite high taxation that restricts the labor supply. In this comparison, U.S. labor market flexibility is seen as the result of low unemployment benefits, low firings costs, and low labor taxation, leading to a more dynamic labor market that can more easily adjust to economic shocks. The exploding cost of unemployment along with public pressure to reduce the high levels of unemployment eventually triggered the largest social policy reform in post-war Germany (Wunsch, 2005; see also Box 4.2). This reform agenda limited the duration and level of unemployment benefits and aimed to improve job creation. As argued in this chapter, these reforms led to a labor market performance in Germany during the recent financial crisis that was quite different from its historical performance as well as from that in other economies.

RECENT DEVELOPMENTS German employment, measured as the number of employed individuals, has performed remarkably during the crisis. Employment decreased only by a moderate 0.6 percent from peak (2008:Q4) to trough (2009:Q4). By contrast, the cumulative peak-to-trough decline in real GDP (2008:Q1–2009:Q1) reached over 6.6 percent. The change in employment is one of the smallest among advanced economies, even compared with those that experienced smaller GDP declines. A comparison with the United States is particularly telling: despite a smaller decline in real GDP, U.S. employment was about 6 percent lower at the trough, a larger decline than that in output (Figure 4.3). These differences were reflected also in terms of recovery—by end-2010, the United States had nearly recovered its precrisis level of GDP, but still remained about 4 percent below its pre-crisis employment level. Conversely, in Germany employment has moved back to precrisis levels quickly, but with output remaining more sluggish (Figure 4.4). Germany’s experience stands out not only in cross-country comparison, but also from the perspective of its historical business cycle dynamics. Despite the unique nature of the financial crisis—both in terms of magnitude and in terms of its global reach—most countries’ labor markets responded in a fashion consistent with historical dynamics. Germany was among only a small group of countries whose labor markets responded so strikingly differently. Figure 4.5 shows actual versus predicted unemployment patterns, based on the cross-country analysis in the 2010 World Economic Outlook (IMF, 2010, ch. 3). Most countries experienced increases in unemployment either close to their predicted values or, in many cases, substantially larger. By contrast, the increase in unemployment in Germany fell substantially short of its predicted value, unique among all countries in the sample.

©International Monetary Fund. Not for Redistribution Schindler 81

18 Real GDP Total employment 16 14 12 10 8 6 4 2 0

Italy Japan Spain IrelandFinland Greece Austria France Canada Germany Portugal Netherlands Switzerland United States United Kingdom

Figure 4.3 Peak-to-Trough Declines in Output and Employment (Percentage points) Sources: IMF, World Economic Outlook; and IMF staff estimates. Note: Calculated as the percent-difference between peak and trough during 2007:Q1–2010:Q3.

101 Germany 100 Sweden 99 United Kingdom 98 France Japan 97 Denmark

96 Finland USA 95

Employment 94 93 92 91 GIPS 90 90 91 92 93 94 95 96 97 98 99 100 101 Real GDP

Figure 4.4 Recovery to Pre-Crisis Levels (2010:Q4, 2008:Q2 = 100) Sources: IMF, World Economic Outlook; and IMF staff estimates. Note: GIPS: Greece, Italy, Portugal, and Spain.

An alternative version of Okun’s law can provide additional insights into the German labor market dynamics (Box 4.1). The past pattern between employment and growth would have predicted a much larger employment decline than observed, given the output loss, on the order of more than one percent (Figure 4.6). In contrast to employment levels, total hours worked declined substantially, by a cumulative 4.4 percent between their pre-crisis peak in 2008:Q2 and their trough in 2009:Q2. When re-estimating a modified Okun’s relationship between economic growth and total hours worked, the peak-to-trough decline implied by

©International Monetary Fund. Not for Redistribution 82 What Does the Crisis Tell Us about the German Labor Market?

8.0 7.0 Actual change Predicted change 6.0 5.0 4.0 3.0 2.0 1.0 0.0 −1.0

Italy USA Japan Spain France GreeceAustria Finland Ireland Norway BelgiumPortugal Sweden Canada Germany Denmark SwitzerlandNetherlands New Zealand United Kingdom

Figure 4.5 Actual versus Predicted Change in Unemployment (Percentage points) Source: IMF, World Economic Outlook (WEO). Note: Based on figure 3.8 in IMF (2010). The predicted changes in unemployment are based on Okun’s law specifications that include additional control variables, including measures of financial, housing and equity market stress.

40.4 Actual employment Predicted

40.2

40.0

39.8

39.6

39.4

2007:Q12007:Q22007:Q32007:Q42008:Q12008:Q22008:Q32008:Q42009:Q12009:Q22009:Q32009:Q42009:Q12010:Q22010:Q3

Figure 4.6 Germany: Employment During the Crisis (Millions) Sources: Deutsche Bundesbank; and IMF staff estimates. Note: The predicted path is based on a regression of total employment on contemporaneous real GDP growth during 1991:Q1–2008:Q1. The estimated coefficient is used to predict employment from 2008:Q2 onwards.

this modification is about 3.9 percent (Figure 4.7). Interpreted in this sense, the labor market response has been less out of line with historical patterns than the employment dynamics suggest, at least during the first stage of the crisis (i.e., from onset to trough). Notably, however, the dynamics in hours and (to a smaller extent) employment were more positive than predicted during the second, recovery stage. This is consistent with a continuation of the longer-term downward trend in unemployment. As argued below, that trend would require

©International Monetary Fund. Not for Redistribution Schindler 83

BOX 4.1 Okun’s Law

Unemployment typically increases when economic growth slows down. Okun (1962) first documented this casual observation, which became known as “Okun’s law,” by postulating various equations that formalized this relationship, including a gap version, which relates the output gap to the deviation of actual unemployment from its natural rate (see also Abel and Bernanke, 2005), and the more commonly used changes version, which relates output growth to changes in the unemployment rate, e.g.:

'Y/Y = α + β'u

where Y is real GDP and u is the unemployment rate. (Different versions of Okun’s law have been estimated in a large literature focusing on the relationship between unemployment and output. See, e.g., Knotek, 2007, and the references therein.) The coefficient β (“Okun’s coefficient”) can be interpreted as the (semi-)elasticity of output growth with respect to 1 percentage changes in the unemployment rate. Okun (1962) estimated a value of β | 3 /3 for the United States during 1947–1960, while later studies have typically found lower values on the order of 2. While Okun’s law has received much empirical support, different versions of it are appropriate in different contexts. In particular, in this paper an alternative version is estimated, namely, the relationship between changes in total employment and real GDP growth, as well as a second specification that considers the relationship between changes in total hours worked and real GDP growth. These specifications are preferable for a number of reasons. First, the particular interest here is on the distinction between adjustment on the external margin—individuals being hired or laid off—and adjustment on the internal margin—changes in hours worked by each worker. The distinction between these two margins turns out to be important in understanding recent labor market dynamics in Germany, but it cannot be made using the traditional specification of Okun’s law that is focused on unemployment. Secondly, and more broadly, employment is arguably the more preferable object of interest, since the unemployment rate may change over the business cycle, not only due to flows between employment and unemployment, but also because of flows into and out of the labor force. Thus, employment is likely to be a more direct measure of economic well- being than unemployment. That said, the broader result in this paper, that Germany’s historical Okun’s law did not hold during the current crisis, is a finding that is not unique to any specification. See, e.g., Figure 4.1 in this chapter and Chapter 3 in IMF (2010).

employers to “overshoot” in the hours/employment recovery, relative to the predicted dynamics, in order to resume the prior downward trend. The dynamics of employment and hours worked during the crisis suggest that German firms may have switched to new adjustment patterns. Historically, employment and total hours typically moved together at shorter frequencies, while hours per worker remained fairly stable around a secular downward trend (Figure 4.8). That is, traditionally, the employment response to short-term fluctuations has largely been on the extensive margin (employment) and less on the intensive margin (hours per worker). In contrast, during this crisis, the adjustment has mainly been on the intensive margin, that is, in hours worked, both during the period of the sharp output fall and during the initial recovery (implying that hours per worker and total hours

©International Monetary Fund. Not for Redistribution 84 What Does the Crisis Tell Us about the German Labor Market?

14.5 Actual hours Predicted 14.4 14.3 14.2 14.1 14.0 13.9 13.8 13.7 13.6 13.5

2007:Q12007:Q22007:Q32007:Q42008:Q12008:Q22008:Q32008:Q42009:Q12009:Q22009:Q32009:Q42010:Q12010:Q22010:Q3 Figure 4.7 Germany: Hours Worked during the Crisis Sources: Deutsche Bundesbank; and IMF staff estimates. Note: The predicted path is based on a regression of total hours worked (on contemporaneous real GDP growth) during 1991:Q1–2008:Q1. The estimated coefficient is used to predict thours from 2008:Q2 onwards.

400 15.5 Hours worked per worker, left axis 390 Total hours worked (in billions), right axis 380 15 370 14.5 360 350 14 340

330 13.5 320 310 13 1991:Q1 1992:Q1 1993:Q1 1994:Q1 1995:Q1 1996:Q1 1997:Q1 1998:Q1 1999:Q1 2000:Q1 2001:Q1 2002:Q1 2003:Q1 2004:Q1 2005:Q1 2006:Q1 2007:Q1 2008:Q1 2009:Q1 2010:Q1 Figure 4.8 Germany: Hours Worked Sources: Federal Statistical Office; and Deutsche Bundesbank.

moved in sync).2 This crucial change in labor dynamics reflected a number of factors, outlined in what follows. The many factors include increased flexibility in hourly adjustments, the temporary nature of the shock (together with the fact that it is less costly to increase hours per worker after a crisis rather than to re-hire workers), and lastly, emerging signs of a shortage of skilled workers.

2 More recently, the continued labor market recovery appears to have started to shift back to the extensive margin, with the increase in total hours worked driven, once again, by employment gains.

©International Monetary Fund. Not for Redistribution Schindler 85

UNDERSTANDING GERMAN LABOR MARKET DYNAMICS To understand Germany’s labor market dynamics, it is useful to view labor market outcomes as a function of institutions, policies, and shocks. Changes in each of these values, along with the interactions between them, shape how labor markets evolve during a crisis. The following discussion brings together a variety of possible factors behind the observed labor market response in each of these categories.

Hartz IV Reforms: Secular Trends Over the past decade, Germany has undertaken a variety of reforms and policy measures that have significantly altered its labor market institutions. A key package has been the so-called Hartz IV reforms, which took effect in 2005 and contained a number of features (Box 3.2). The reforms streamlined and reduced the generosity of social transfers, including benefits associated with unemployment insurance (UI) and welfare payments, and also deregulated temporary employment. A key effect was to lower unemployed individuals’ reservation wages and thus increase the flows from unemployment into employment. The reforms came after a cyclical acceleration of the longer-term trend increase in unemployment—starting from around 5 percent around the time of German unification, the unemployment rate rose to a peak of over 11 percent in 2005. Although GDP growth had the ability to lower unemployment at shorter (cyclical) frequencies, the structural increase dominated (Figures 4.9 and 4.10). Near zero real output growth during 2001 through 2003 then provided an additional “boost” to unemployment and helped create the political economy that laid the groundwork for the Hartz reforms to take place. Model simulations suggest that the Hartz reforms, especially those in stage IV, contributed strongly to reducing structural unemployment. Analyzing and assess- ing the payoffs from a single reform episode is empirically difficult, as the reform’s impact cannot easily be disentangled from the contributions of simultaneous changes in other factors. In such cases, a theoretical framework can provide guidance on the isolated impact of reforms, since it allows one to construct a counterfactual prediction of how the unemployment dynamics would have evolved post-reform, holding other factors constant. In the context of labor market reforms and their impact on employment and unemployment, the search framework is ideally suited for this task. Labor search models center on the flows that occur between unemployment and employment and the various factors that may affect these flows.3 In particular, simple search models involve variables regarding the payoff from working (in particular, the wage) as well as the payoff from being unemployed (the monetary

3 More complex models may include also the flows from outside the labor force into employment or unemployment, and vice versa. For tractability, the model considered here abstracts from these.

©International Monetary Fund. Not for Redistribution 86 What Does the Crisis Tell Us about the German Labor Market?

10.0 Real GDP (Percent change over 4 quarters 3.0 ago), left axis 8.0 Unemployment rate (ILO, percentage point 2.0 6.0 change over 4 quarters ago), right axis

4.0 1.0 2.0

0.0 0.0

−2.0 −1.0 −4.0

−6.0 −2.0 −8.0

−10.0 −3.0 1991:Q1 1992:Q1 1993:Q1 1994:Q1 1995:Q1 1996:Q1 1997:Q1 1998:Q1 1999:Q1 2000:Q1 2001:Q1 2002:Q1 2003:Q1 2004:Q1 2005:Q1 2006:Q1 2007:Q1 2008:Q1 2009:Q1 2010:Q1 Figure 4.9 Germany: Cyclical Changes in Output and Unemployment Sources: Federal Statistical Office; and Deutsche Bundesbank.

Real GDP (Seasonally and work day adjusted, 115.0 2000 = 100), left axis 12.0 Unemployment rate (ILO, percent), right axis 110.0 10.0 105.0

100.0 8.0

95.0 6.0 90.0 4.0 85.0

80.0 2.0 2000:Q1 2006:Q1 1991:Q1 1992:Q1 1993:Q1 1994:Q1 1995:Q1 1996:Q1 1997:Q1 1998:Q1 1999:Q1 2001:Q1 2002:Q1 2003:Q1 2004:Q1 2005:Q1 2007:Q1 2008:Q1 2009:Q1 2010:Q1 Figure 4.10 Secular Trends in Output and Unemployment Sources: Federal Statistical Office; and Deutsche Bundesbank.

and nonmonetary value derived from being unemployed). These factors directly impact the extent to which an unemployed individual will actively search, as well as which job opportunities might be taken up. Another important factor for labor market outcomes is the matching efficiency, that is, the ease with which firms encounter the right candidates and unemployed individuals seeking work find the

©International Monetary Fund. Not for Redistribution Schindler 87

14 Unemployment rate (actual) 13 Simulated unemployment (USS = 5.25 percent) Simulated unemployment (USS = 6.0 percent) 12 Simulated unemployment (USS = 7.0 percent) 11 Simulated unemployment (USS = 8.1 percent (historical average))

5 2000:Q1 2000:Q3 2001:Q1 2001:Q3 2002:Q1 2002:Q3 2003:Q1 2003:Q3 2004:Q1 2004:Q3 2005:Q1 2005:Q3 2006:Q1 2006:Q3 2007:Q1 2007:Q3 2008:Q1 2008:Q3 2009:Q1 2009:Q3 2010:Q1 2010:Q3 Figure 4.11 Actual and Simulated Unemployment (Percent) Sources: Deutsche Bundesbank; and IMF staff estimates. Note: Hypothetical paths are model-based transition paths under the assumption that the Hartz IV reforms lowered the structural (steady-state) unemployment rate (USS) to various levels. See text for details. right jobs. The appendix outlines a reduced-form version of a search model, with a focus on the implied transition dynamics.4 Model simulations suggest that the decline in unemployment observed since 2005 is likely to continue. From an average of about 8.1 percent during 1991– 2005 and a peak of 10.6 percent in 2005, the unemployment rate subsequently started a downward trend that was interrupted only briefly at the height of the financial crisis. Most recently, the unemployment rate, at just under 6 percent in 2011:Q2, has reached a level not seen in nearly two decades. The model suggests that an increase in the rate at which workers encounter the right jobs can explain much of the precrisis trend in the unemployment rate. More specifically, the model as described by equation (1) in the appendix contains two key parameters: the rate at which unemployed individuals find work, α, and the rate at which workers lose employment, δ. This equation describes the unemployment dynamics, given that current unemployment is equal to last period’s unemployment minus the share of unemployment that has found employment plus the share of the employed that has lost employment.5 Different parameterizations of the equation imply different structural (steady-state) unemployment rates. To the extent that the Hartz reforms raised the job-finding rate, these reforms imply a lower steady-state unemployment rate (Figure 4.11).

4 An important feature of this reduced-form model is that one can write down an explicit search- model which, for the right set of parameters, can rationalize the reduced form model used here. 5 As previously mentioned, these simplifying assumptions include ignoring movements into and out of the labor force, as well as any heterogeneity in the workforce, whereby different worker groups may have potentially different job finding/separation rates. Another important assumption is that of constant parameters over time (with the exception of discrete changes at the time of a shock).

©International Monetary Fund. Not for Redistribution 88 What Does the Crisis Tell Us about the German Labor Market?

Starting from the pre-Hartz IV unemployment rate of about 10½ percent, and assigning various parameter values to the post-reform value of α, equation (1) can also be used to trace out the simulated unemployment transitions to a hypothetical new steady-state (Figure 4.9).6 From the peak in 2005 through the onset of the crisis, the simulated path based on a 5.25 percent long-term unemployment rate gives an almost perfect fit with the post-2005 data. In this view, the crisis temporarily halted the trend decline, but employment quickly recovered and is likely to continue its downward trend.7 The simulations based on a 5.25 percent long-term rate imply that the job-finding rate would be about 50 percent higher than prior to Hartz IV, an implication that is broadly supported by other research (Lam, 2011). The interpretation of the post-2005 path as a transition to a lower structural unemployment rate also affects the interpretation of the observed labor market response. If the downward trend since 2005 is indeed a transition to a new and lower steady-state unemployment rate, rather than a low-frequency cyclical variation around the previous unemployment rate of more than 8 percent, then the unemployment dynamics during the crisis are best assessed relative to the counterfactual transition path rather than relative to precrisis levels of unemployment. Thus, compared to what unemployment might have been in the absence of the crisis, unemployment at its crisis peak in 2009:Q3 had increased by over one percentage point relative to the counterfactual—this is more than twice the change calculated as the difference between the precrisis level of 7.2 percent and the crisis peak of 7.6 percent. In this reading, then, the German labor market response has not been quite as mild as typically understood.8

Hourly Flexibility Measures Some reforms, prior to and during the crisis, were targeted at hourly flexibility. The most prominent such measure included extensions to the German short-time scheme (Kurzarbeit), which subsidized workers and firms to reduce hours worked. This measure was taken up extensively by workers and firms. However, it was only a second or third step for many firms. In years prior, employers and unions had agreed on a number of measures that facilitated reductions in the work week (with commensurate income reductions) and also helped firms to smooth hours per worker over time through work-time accounts (Arbeitszeitkonten) (Box 4.2). Many workers had accumulated large surpluses in their work-time accounts (i.e., worked more than their regular work weeks), and thus firms had substantial buf- fers to adjust labor input by exhausting these measures before turning to short- time subsidies (see Sachverständigenrat, 2009).

6 See the appendix for details on the construction of the hypothetical paths. 7 However, the simulations are suggestive that unemployment may have shifted to a different trend line. 8 As a corollary, employment growth could also have been stronger absent the crisis, suggesting that the above- noted mild employment response may underestimate the true damage from the crisis.

©International Monetary Fund. Not for Redistribution Schindler 89

BOX 4.2 Recent Changes in German Labor Market Institutions

Recent reforms in the German labor market were mainly in two broad areas: (i) the Hartz reforms, which, broadly speaking, facilitated movements from unemployment to employment, and (ii) measures to increase hourly flexibility, facilitating firms’ adjustment of labor input without necessarily changing the number of employees. This box outlines some of the key features of the various reforms. i. The Hartz reforms were a sequence of reform packages aimed at implementing the 2002 proposals by the committee on “Modern Services in the Labor Market.”a Their overall intent was to facilitate the flow from unemployment to employment, both by lowering benefits during unemployment (Hartz IV) and by adding measures to improve intermediation (Hartz I through III). More specifically: – Key aims of the first installments, the Hartz I and II reforms (both effective January 2003) were to increase and facilitate temporary employment (e.g., through personnel service agencies) and to raise the incentives for individuals receiving social benefits to engage in productive work (e.g., the Ich-AG and job centers). – The Hartz III package, effective January 2004, restructured the federal labor agency. – The last, and arguably most important, stage came with the Hartz IV measures, effective January 2005. This fourth stage shortened the eligibility for unemployment insurance benefits (Arbeitslosengeld) to 18 months from the previous 24 months and set subsequent unemployment benefits (Arbeitslosenhilfe) at the substantially lower level of social welfare benefits (Sozialhilfe). ii. A second set of changes in the labor market regards hourly flexibility. The following list summarizes some of the key measures: – In the early 2000s, an increasing number of sectors and industries, through agreements between workers and employers, introduced opening clauses (Öffnungsklauseln) that allowed them to deviate from collectively bargained work arrangements, e.g., to reduce worktime without compensation, or to reduce pay in exchange for avoiding layoffs. – Firms have also increasingly taken advantage of worktime accounts (Arbeitszeitkonten). The main feature of these is to allow firms to vary workers’ hours worked over time, reducing the need for overtime pay and minimizing workers’ income fluctuations. The worktime account ensures that hourly fluctuations balance out inter-temporally. – Short-time subsidies (Kurzarbeit) by the government have been a feature of German labor market institutions for nearly a century. They facilitate work-time reductions by partially compensating workers for up to 67 percent of their lost income. The subsidy was made more generous in 2008–09 through a temporarily longer maximum duration (24 months) and through increased coverage of firms’ social contributions.

a The committee was chaired by Peter Hartz, then-member of the executive board of Volkswagen Aktiengesellschaft in charge of human resources and personnel management.

The Nature of the Shock Germany experienced a different shock than, for example, the United States. While the German economy was affected strongly and real GDP declined by more than in the United States, the shock was largely externally driven, channeled to Germany through a decline in export demand. Given Germany’s strong export dependence, the impact on GDP was severe, but at the same time it was short- lived—exports recovered almost as quickly as they had deteriorated. Partly as a

©International Monetary Fund. Not for Redistribution 90 What Does the Crisis Tell Us about the German Labor Market?

result, Germany’s growth potential was also little affected. (See Chapter 2). The nature of the shock differed substantially from that in the United States, which suffered a more domestic and structural shock to its economy.

Interactions between Institutions, Policies and the Shock Firms’ expectations of a temporary shock only partially explain the strong reliance on hourly adjustment. Although the downturn has indeed proven temporary, with a strong recovery that started in mid-2009, the cumulative loss has been large—for example, compared to precrisis output levels, despite its robust recovery Germany still lags behind the United States, while the latter has suffered a large and persistent increase in unemployment.9 The labor market dynamics may reflect the still-high layoff costs. While the level of employment protection legislation (EPL) has not markedly changed in the recent past10 or during the crisis, and so cannot by itself explain the deviation in Germany’s labor market response from its past behavior, it is possible that the increased hourly flexibility has tilted firms’ cost-benefit analysis, making reductions in hours per worker a less costly alternative to employment reductions.11 Labor supply constraints in the form of a shortage of skilled workers may also have contributed, but the evidence is not fully conclusive. A shortage of skilled workers is frequently cited by firms, academics, and policy makers alike as being at the root of firms’ labor hoarding behavior during the crisis. Consistent with this, a duality has started to emerge in the German labor market, whereby firms have largely relied on extensive labor adjustment among temporary workers, who

9 The notion that firms rationally responded to a temporary shock by hoarding labor does not require the assumption of perfect foresight—e.g., if firms use some form of Bayesian updating, then the low level of external demand at the trough of the crisis would not have been perceived as the new equilibrium without it persisting for a sufficiently long time. 10 Germany continues to rank high according to the OECD’s employment protection legislation (EPL) indicators, especially on regular employment (on temporary employment, Germany was below the OECD average in 2008, the latest available data point). Consequently, the OECD has repeatedly called on Germany for further reform in this area (OECD, 2011). In addition, Berger and Neugart (2012) point to an additional cost of reducing employment in the form of uncertainty over labor court decisions. In terms of labor market implications, because EPL tends to reduce both job destruction and job creation, the impact on the level of employment is conceptually unclear, reflected in similarly ambiguous empirical findings (Garibaldi and Mauro, 2002, versus Takizawa, 2003). But there is a broader consensus that high EPL inhibits labor market dynamism and its ability to respond to economic shocks. See Barone (2011) and Schindler (2009) for reviews of related literatures. In addition, Boeri (1999) argues that the asymmetric type of EPL, as increasingly observed in Germany (i.e., regular versus temporary employment) can lead to labor market dualism. 11 That is, the same level of employment protection legislation may have a larger impact on allocations when firms have new ways of avoiding layoff costs, as in this case through more hourly flexibility. See Ljunqvist and Sargent (1998) for a similar reasoning on the interactions of institutions and shocks.

©International Monetary Fund. Not for Redistribution Schindler 91 are typically less skilled, while relying on more intensive adjustment for their more skilled core workforce.12 However, despite much anecdotal evidence, labor market outcomes do not fully bear out the signs of a true shortage: Brenke (2010) has noted that such a shortage should put upward pressure on their relative wages, something that has not been observed. Also, university enrollment rates in engineering and the natural sciences remain at a high level and have in recent years increased substantially, suggesting that supply is relatively strong.13 While these statistics suggest that excess demand for skilled labor in the German labor market may not be as broad- based and severe as some may fear, there is little doubt that demographic pressures will pose problems in the medium term. Matching inefficiencies appear to have diminished. Even without shortages of skilled labor, firms may opt for “labor hoarding” if they find it difficult to rehire the workers they need in an upswing. The relationship between vacancies and unemployment over time (the so-called Beveridge curve, first discussed by Dow and Dicks-Mireaux, 1958) is a frequently used gauge for a country’s matching efficiency.14 It is especially insightful in the context of search models—in such models, labor supply (unemployed workers) and labor demand (firms’ vacant positions) are brought together through a matching technology, or matching function. A more efficient matching technology then implies that a given level of vacancies translates into a lower unemployment rate than would be the case with a less efficient matching technology. Graphically, shifts in the Beveridge curve toward the origin suggest gains matching efficiency, while movements along the curve are the result of equilibrium dynamics for a constant matching technology. The plot for Germany suggests that around 2007, the matching process started to improve, with the Beveridge curve now at a substantially lower level (i.e., a shift toward the origin) than pre-2007 (Figure 4.12). This apparent improvement in matching efficiency is consistent with the previously noted faster unemployment-to-employment transition associated with the Hartz reforms, including, importantly, the measures taken during the first two stages that facilitated the development of atypical employment. All else equal, a higher matching efficiency should increase firms’ willingness to shed labor, on the expectation that finding a good match in the future will be relatively easy. While the shift in the Beveridge curve makes it at first sight more difficult to understand the stable employment during the crisis, the aggregate data hide the fact that the Hartz reforms mostly added labor supply on the lower end of the income and skill distribution. That addition is consistent with the fact that employment losses were larger among low-wage segments, suggesting that

12 Employment of temporary workers declined by 2 percent in Germany during 2009, compared with an increase of 0.4 percent in total employment (IMF, 2010, Box 3.1). 13 However, enrollment rates in these fields were broadly flat during 2004–2008, suggesting that in the coming years, graduation rates may not keep up with demand of a growing economy. 14 Vacancies and unemployment are negatively correlated, since strong hiring demand by firms (high level of vacancies) helps reduce unemployment.

©International Monetary Fund. Not for Redistribution 92 What Does the Crisis Tell Us about the German Labor Market?

9 2000–07 8

7 2008–present 6 150 200 250 300 350 400 450

Figure 4.12 Beveridge Curve Source: Deutsche Bundesbank. Note: The Beveridge curve plots monthly unemployment rates (in percent, sa) against the monthly number of job vacancies (in thousands, sa).

improvements in employment flexibility and matching efficiency were more pronounced in those segments.

CONCLUSION The apparent puzzle of the German labor market during the crisis can be explained by a combination of factors. The German economy suffered a demand shock, which firms presumably expected to eventually reverse. Hence, firms adjusted mainly by taking advantage of options to reduce the hours worked, keeping employment levels relatively high (i.e., relative to that predicted by Okun’s law). At the same time, the German unemployment rate had been on a downward path since about 2005, most likely reflecting the Hartz IV reforms. This also had the effects of dampening the decline in employment and helping the employment recovery. While authors in the existing literature place emphasis on different factors (Box 4.3), most agree that the likely explanation involves some combination of hourly flexibility measures, including work-time accounts and short-time subsidies, low precrisis hiring, and long-standing wage moderation. While these views are consistent with the interpretation put forth in this paper, the emphasis here is different: namely, in this analysis most of the credit belongs to the Hartz reforms, which have put Germany in a position to deal well with a temporary demand shock. This still leaves ample room for further reforms. The increased flexibility of the lower-wage segment has introduced an uneven distribution of employment and wage risk: more secure high-skill and high-wage employment along with less secure low-wage work. Also, while job creation has been made easier, rigidities still remain on the job termination side, resulting in part from

©International Monetary Fund. Not for Redistribution Schindler 93

BOX 4.3 The Literature on the “German Employment Miracle”

The academic literature on the German performance during the crisis is still small, unsurprisingly so given the proximity of events. No clear consensus has yet emerged on the key drivers of the German performance. Burda and Hunt (2011) cite three main factors as the reasons for the German “labor market miracle:” relatively low precrisis hiring, due to a lack of confidence by firms in the robustness of the precrisis boom, provided a buffer during the crisis and accounted for about a third of the missing employment decline; the secular trend in wage moderation, which tends to raise employment, is seen as explaining 10 percent of the employment decline that would have otherwise been expected; and work-time accounts are seen as the third factor. Burda and Hunt do not discuss the possibility that the low precrisis hiring and accumulation of large surpluses in the work-time accounts may have similar sources and thus may not be truly separate explanations of the German labor market performance. They also do not fully explain how low precrisis hiring is consistent with the fairly strong employment gains that were made almost immediately after the initial crisis shock had worn off. Lastly, just under half of the unexplained employment strength is explained in their analysis. Other authors emphasize different reasons for Germany’s labor market performance. Gartner and Merkl (2011) also emphasize precrisis wage moderation as the main source of the crisis performance, since it had in their view induced a permanent labor supply shock. It is not obvious, however, how a permanent shift would impact the employment dynamics in a cyclical downturn. It is also possible that this link between wage moderation and reduced employment losses during the crisis reverses causality—if, as argued in this chapter, the Hartz reforms were the underlying reason for the expansion in employment, then it is the “Hartz supply shock” that weakened workers’ bargaining positions and put further moderating pressure on wages, rather than wage moderation causing employment. Boysen-Hogrefe and Groll (2010) and Gartner and Klinger (2010), in turn, place stronger emphasis on work-time accounts as the key source of the robust employment dynamics during the crisis. Fahr and Sunde (2009) and Klinger and Rothe (2010) take a view that is most closely aligned with that in this chapter. Namely, they argue that the Hartz reforms raised the efficiency of the matching function increased, consistent with Gartner and Klinger’s (2010) observation of a continued shift in the Beveridge curve.

still-high employment protection and uncertainties related to labor court proceedings. Lastly, demographic pressures in the medium term will necessitate broader measures, including in education and immigration, to moderate the emergence of bottlenecks. Policy conclusions for other countries are to be drawn cautiously. As argued in this note, the German crisis performance reflected the type of shock, crisis policies, and initial conditions at the time of the shock. The latter, notably, reflected reforms initiated several years prior to the crisis, which put German unemployment on a declining trend that mitigated the increase during the crisis. A country with differences in any of these aspects may well require a different policy response, so blindly applying measures such as the Kurzarbeit extensions could well have adverse effects if, for example, the shock were more structural in nature. One broader lesson, however, is that the “right” reforms are likely to eventually pay off.

©International Monetary Fund. Not for Redistribution 94 What Does the Crisis Tell Us about the German Labor Market?

REFERENCES Abel, Andrew B., and Ben Bernanke, 2005, Macroeconomics, 5th edition. Addison-Wesley. Barone, Andrea, 2011, “Employment Protection Legislation: A Critical Review of the Literature.” Published online at http://jpkc.ojc.zj.cn/ldfx/UploadFiles/2010928214832623. pdf, downloaded October 18, 2011. Berger, Helge, and Michael Neugart, 2012, “How German Labor Courts Decide—An Econometric Case Study,” German Economic Review, Vol. 13, pp. 56–70. Boeri, Tito, 1999, “Enforcement of Employment Security Regulations, On-the-Job Search and Unemployment Duration,” European Economic Review, Vol. 43, pp. 65–89. Boysen-Hogrefe, Jens, and Dominik Groll, 2010, “The German Labour Market Miracle,” National Institute Economic Review, Vol. 214, pp. R38–R50. Brenke, Karl, 2010, “Fachkräftemangel kurzfristig noch nicht in Sicht,“ DIW Berlin Wochenbericht No. 46 (2010). Burda, Michael, and Jennifer Hunt, 2011, “What Explains the German Labor Market Miracle in the Great Recession?,” NBER Working Paper No. 17187 (Cambridge: National Bureau of Economic Research). Dow, J. C. R., and L. Dicks-Mireaux, 1958, “The Excess Demand for Labour: A Study of Conditions in Great Britain, 1946–1956,” Oxford Economic Papers, No. 10, pp. 1–33. The Economist, 1999, “The Sick Man of the Euro,” June 3rd (downloaded from http://www. economist.com/node/209559). Fahr, René, and Uwe Sunde, 2009, “Did the Hartz Reforms Speed Up the Matching Process? A Macro-Evaluation Using Empirical Matching Functions,” German Economic Review, Vol. 10, pp. 284–316. Garibaldi, Pietro, and Paolo Mauro, 2002, “Anatomy of Employment Growth,” Economic Policy, Vol. 17, pp. 67–113. Gartner, Hermann, and Sabine Klinger, 2010, “Verbesserte Institutionen für den Arbeitsmarkt in der Wirtschaftskrise,” Wirtschaftsdienst, Vol. 90, pp. 728–34. Gartner, Hermann, and Christian Merkl, 2011, “Die ökonomische Basis des Arbeits- marktwunders,” http://www.oekonomenstimme.org/artikel/2011/03/die-oekonomische-basis- des-arbeitsmarktwunders/. IMF (International Monetary Fund), 2010, World Economic Outlook: Rebalancing Growth, April (Washington, DC: International Monetary Fund). Klinger, Sabine, and Thomas Rothe, 2010, “The Impact of Labour Market Reforms and Economic Performance on the Matching of Short-Term and Long-Term Unemployed,” IAB Discussion Paper No. 13/2010 (Nuremburg: Institute for Employment Research (IAB) of the German Federal Employment Agency (BA)). Knotek, II, Edward S., 2007, “How Useful is Okun’s Law?,” Federal Reserve Bank of Kansas City Economic Review, Vol. 4, pp. 73−103. Lam, W. Raphael, 2011, “Does Labor Market Flexibility Explain Unemployment Dynamics in the Great Recession?,” IMF Working Paper, forthcoming (Washington, DC: International Monetary Fund). Ljungqvist, Lars, and Thomas Sargent, 1998, “The European Unemployment Dilemma,” Journal of Political Economy, Vol. 106, pp. 514−50. OECD (Organisation for Economic Co-operation and Development), 2011, Economic Policy Reforms 2011: Going for Growth (Paris: OECD). Okun, Arthur M., 1962, “Potential GNP: Its Measurement and Significance,” Proceedings of the Business and Economics Statistics Section, American Statistical Association, pp. 98–104. Pissarides, Christopher A., 2000, Equilibrium Unemployment Theory, 2nd edition (Cambridge, Massachusetts: The MIT Press). Prescott, Edward C., 2004, “Why Do Americans Work So Much More Than Europeans?” Federal Reserve Bank of Minneapolis Quarterly Review, Vol. 28, pp. 2–13. Rogerson, Richard, 2010, The Impact of Labor Taxes on Labor Supply: An International Perspective (Washington, DC: AEI Press).

©International Monetary Fund. Not for Redistribution Schindler 95

Sachverständigenrat für Wirtschaft (Sachverständigenrat), 2009, “Die Zukunft nicht aufs Spiel setzen,” Jahresgutachen 2009/10. Schindler, Martin, 2009, “The Italian Labor Market: Recent Trends, Institutions, and Reform Options,” IMF Working Paper No. 09/47 (Washington, DC: International Monetary Fund. Siebert, Horst, 1997, “Labor Market Rigidities: At the Root of Unemployment in Europe,” Journal of Economic Literature, Vol. 11, pp. 37−54. ———, 2003, “The Failure of the German Labor Market,” Kiel Working Paper No. 1169 (Kiel: Kiel Institute for World Economics). Takizawa, Hajime, 2003, “Job-Specific Investment and the Cost of Dismissal Restrictions—The Case of Portugal,” IMF Working Paper No. 03/75 (Washington, DC: International Monetary Fund). Wunsch, Conny, 2005, “Labour Market Policy in Germany: Institutions, Instruments and Reforms since Unification,” University of St. Gallen, Department of Economics, Discussion Paper Nr. 2005–06.

©International Monetary Fund. Not for Redistribution 96 What Does the Crisis Tell Us about the German Labor Market?

APPENDIX A simple search model framework can help in understanding the impact of the Hartz IV reforms on unemployment. In labor markets with search frictions, transitions from unemployment to employment are not instantaneous. Ignoring changes to the labor force, unemployment can thus be seen to evolve over time according to:

ut + 1 = ut(1 − α) + (1 − ut) δ (1)

where ut is unemployment at time t and α,δ are the probabilities (hazard rates) of exiting unemployment and employment, respectively. This equation also implies that steady-state unemployment can be expressed as uSS = δ / (δ + α ). (2) The Hartz IV reforms have likely lowered structural unemployment through a higher job finding rate. The pre-Hartz IV average unemployment rate was about 8.1 percent. Lam (2011) estimates the German job destruction rate at about 0.5 percent. These two values imply a pre-2005 average job finding rate of about 6.1 percent.15 Assuming δ to have remained unchanged since 2005, the steady-state equation provides a one-to-one correspondence between estimates of the (new) long-term steady-state and the implied (new) job finding rate α.16 The figure in the main text plots the resulting transition paths for a variety of possible values of the new structural unemployment rate. See also Fahr and Sunde (2009), Klinger and Rothe (2010), and Gartner and Klinger (2011) for a similar view of events.

15 Lam (2011) estimates job finding and destruction rates for 1970–2009 and for 1996–2009 and finds little difference for the different time periods (for the latter period, he estimates δ = 0.6 percent). The implied value for α corresponds to Lam’s estimate who also finds a value of also 6.1 percent during 1970–2009. 16 For the purpose of this reduced-form exercise, we take the job arrival and destruction rates as exogenous. In a more general search model, both margins would be endogenous due to changes in individuals’ reservation wages. Generally speaking, a reduction in unemployment benefits would raise job creation and lower job destruction (Pissarides, 2000)—in such models, the key channel is through wage bargaining.

©International Monetary Fund. Not for Redistribution CHAPTER 5 Growth Spillover Dynamics: From Crisis to Recovery

HÉLÈNE POIRSON AND SEBASTIAN WEBER

Can positive growth shocks from the faster-growing countries in Europe spill over to the slower-growing countries, providing useful tailwinds to their recovery process? This study investigates the potential relevance of growth spillovers in the context of the crisis and the recovery process. Based on a Vector Autoregression (VAR) framework, our analysis suggests that the United States and Japan remain the key sources of growth spillovers in this recovery, with France also playing an important role for the European crisis countries. Notwithstanding the current export-led cyclical upswing, Germany generates relatively small outward spillovers compared to other systemic countries, but it likely plays a key role in transmitting and amplifying external growth shocks to the rest of Europe given its more direct exposure to foreign shocks compared to other European countries. Positive spillovers from Spain were important prior to the 2008– 09 crisis, but Spain is generating negative spillovers in this recovery due to its depressed domestic demand. Negative spillovers from the European crisis countries appear limited, consistent with their modest size.

INTRODUCTION Can positive growth impulses from the faster-growing countries in Europe spill over to the slower growing countries, providing useful tailwinds to their recovery process? This analysis investigates the relevance of such potential growth spillovers and seeks to identify the countries most likely to serve as “growth leaders.” In particular, we examine the extent to which Germany’s current upswing may spill over to other countries and accelerate recovery elsewhere. A simple correlation between the lagged quarter-on-quarter output growth rates of Germany and other euro area countries shows an increasing co-movement between Germany and other countries, including the European crisis countries (Greece, Ireland, and Portugal) in the last 20 years. A similar pattern can be

The authors thank Ashoka Mody for insightful comments and valuable suggestions. They also would like to thank Céline Allard, Ansgar Belke, Rupa Duttagupta, Felix Huefner, Irina Tytell, and Francis Vitek, participants in the Germany in an Interconnected World Economy conference (Berlin, 2011) and in the Graduate Institute of International and Development Studies internal seminar (Geneva) for useful discussions and comments; and Susan Becker for excellent research assistance. 97

©International Monetary Fund. Not for Redistribution 98 Growth Spillover Dynamics: From Crisis to Recovery

observed for the correlations of the lagged growth rates of France and Italy with the current GDP growth rate of the rest of the euro area. The correlation of the lagged U.S. growth rate with other countries’ current GDP growth rates has also increased markedly (Figure 5.1). While more synchronized business cycles suggest increased growth spillovers, they do not provide a measure of the spillover effects of individual countries. Moreover, they could also reflect the growing influence of common factors, such as euro area monetary policy, or global factors, such as oil price developments or financial conditions in major financial centers including the United States and the United Kingdom. To disentangle the independent impact of growth shocks in individual countries from the effect of common factors and measure the country-specific spillover effects, we use a structural vector autoregression (SVAR) approach with an identification scheme similar to that proposed in Bayoumi and Swiston (2009). Using this framework, we undertake the following analysis: • We characterize countries by their international growth spillovers, outward and inward. Unlike former studies, which rely on regional aggregates, we use country-level data for a sample of 17 countries (accounting for almost 60 percent of global GDP, at market exchange rates), which includes 11 of the euro area countries (representing 98 percent of euro area output). Focusing on individual members rather than treating the euro area as an aggregate as in earlier studies allows us to shed light on the country-by-country spillovers and inter-linkages that drive growth dynamics. • We describe the dynamics of these spillovers during the recent crisis and recovery. This is done by applying a dynamic growth accounting calculation to the SVAR estimation results to quantify the contribution of individual countries to the recent decline and recovery in output growth. • We quantify the relevance of different channels of transmission of spillovers. We use counterfactual analysis and smaller country–by-country regressions, which include exports as an additional variable to assess the empirical relevance of various transmission channels, including trade and third-country effects (e.g., transmission through a common trade partner). • We analyze the determinants of the spillover size by relating outward spillovers to both the country’s size and the relative importance of the domestic demand contribution to the country’s growth. Our main findings can be summarized as follows: • Confirming earlier results, we find evidence of significant spillover effects from the United States to the rest of the world. Outward spillovers from Germany are found to be surprisingly small (relative to growth impulses emanating from other large systemic countries), and they are largely con- fined to smaller trade partners. • The results suggest that despite the increased correlation of Germany’s growth rate with that of the rest of the euro area, the United States and Japan remain the key sources of growth spillovers in this recovery, with

©International Monetary Fund. Not for Redistribution Poirson and Weber 99

1 Correlation(Country (t-1), EMU) 1 Correlation(Country (t-1), GIP) United States Japan United Kingdom Germany 0.8 France Italy 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0

−0.2 −0.2 1975- 1980- 1985- 1990- 1995- 2000- 2005- 2009- 1975- 1980- 1985- 1990- 1995- 2000- 2005- 2009- 1979 1984 1989 1994 1999 2004 2008 2010 1979 1984 1989 1994 1999 2004 2008 2010

Figure 5.1 Growth Correlation with EMU and GIP Countries over Time: Selected Economies (1975–2010) Source: OECD, IMF staff calculations. Note: EMU: Austria, Belgium, Finland, Greece, Ireland, Netherlands, Portugal, and Spain, weighted by their respective share in the group; GIP: Greece, Ireland, and Portugal.

France also playing an important role for the European crisis countries. Positive spillovers from Spain were important prior to the crisis. However, Spain is generating negative spillovers in this recovery due to its depressed domestic demand. • In line with earlier VAR-based evidence on channels of spillover transmission across regions, the results of a country-by-country estimation of a smaller SVAR model (augmented with exports) suggest that financial and other nontrade channels explain the biggest share of cross-border growth spillovers. However, trade effects are particularly relevant in the case of Germany. • Taking the analysis one step further, we find that countries that generate the largest estimated outward spillovers are the ones where growth is largely driven by autonomous sources of domestic demand. In contrast, countries highly sensitive to external shocks, such as Germany, tend to have a relatively small independent, aggregate impact on other countries. This result, together with the finding that third-country effects play a significant role in the international transmission of shocks, suggests that Germany plays an important role as a transmitter and amplifier of growth shocks originating in other countries. Three main caveats should be mentioned at the outset. All three provide inter- esting avenues for future research but are beyond the scope of this paper. • Countries outside our sample might play an important role in global spillovers, either as recipients or as a source of shocks. For instance, several

©International Monetary Fund. Not for Redistribution 100 Growth Spillover Dynamics: From Crisis to Recovery

Eastern European countries are highly integrated with Germany and have become an integral part of the production chain of German products. For these countries, empirical evidence supports the idea that Germany has played a more prominent role in recent years (see Danninger, 2008). In recent years, China and other large emerging countries probably play an important role as a source of shocks as well, reflecting their size and increased trade linkages.1 • There may be other sources of growth spillovers beyond the business-cycle fluctuations considered in this study. For example, foreign direct investment (FDI) outflows between advanced countries and from advanced to developing countries could generate significant spillovers over time, including through knowledge transfer and employment creation. Quantifying such FDI-related spillovers would require a different approach, one focused on a long-term horizon, since such effects are likely to develop only over time. • The analysis here is backward looking and does not allow for continuous time-varying relationships among the countries. In addition to full sample results, we present estimation results for a more recent sub-sample (since 1993) to account for the possibility of changed responses, in line with rapidly increasing cross-border trade and financial linkages. The results confirm that the potential size of cross-border growth spillovers has generally increased, with the notable exceptions of Japan and Germany, for which we find no evidence of higher outward spillovers on average in the recent period. Finally, it should be noted that this analysis is descriptive. An analysis of welfare implications is beyond the scope of this chapter. The remainder of this chapter reviews the related literature on cross-country spillovers, outlining the main empirical approaches that have been pursued and the key findings; discusses the empirical strategies employed in this study; and high- lights the main findings and provides estimates of individual countries’ relevance for other countries’ growth dynamics. The potential reasons for the relevance of particular countries as a source of spillovers are also discussed.

GROWTH LINKAGES AND SPILLOVERS: RELATED LITERATURE The empirical literature on growth spillovers has dealt with three interrelated questions: What is the size of growth spillovers? Which countries or regions are

1 While data availability constrains the inclusion of China and other emerging countries in the sample, we indirectly attempted to test this hypothesis in an earlier version of the study by including exports to China, to developing Asia, and to the world as additional control variables. Those variables were not statistically significant and the results remained largely unchanged, suggesting that the sample countries already capture the bulk of relevant global demand shocks for the period under consideration (1975–2010).

©International Monetary Fund. Not for Redistribution Poirson and Weber 101 the main sources of growth spillovers? and, What are the main channels of transmission of growth spillovers? Existing empirical studies generally confirm the existence of spillover effects, but the empirical evidence about the size of spillovers is inconclusive, as results are not robust to different samples and specifications. Previous studies frequently divide the world into regions,2 the euro area being one of them, and examine spillovers from large advanced countries to these regions.3 The general finding of these studies is that the United States is the main source of growth spillovers. Relatively few studies have examined the channels through which growth shocks are transmitted to the other regions and countries. Existing findings vary across studies, with simulation-based results suggesting a bigger role for the trade channel—perhaps due to the difficulty of empirically modeling asset price spillovers or confidence channels—while VAR analyses, which impose less structure on the interlinkages, point to the relative importance of financial and other nontrade channels. The finding that international spillovers are relatively small under standard transmission channels was first established by Helbling and others (2007). Their study finds a limited extent of U.S. growth spillovers into other regions—excluding the euro area and Japan—and even smaller effects of spillovers from the euro area or Japan, when controlling for possible channels of transmission including commodity prices (terms of trade) and financial conditions (Libor interest rate). The results obtained are similar using three alternative approaches (simple panel regressions, a more sophisticated dynamic analysis, and model simulations). The estimated spillovers are moderate in magnitude: the results from annual panel regressions for 130 countries over 1970–2005 suggest that a 1 percentage point decline in U.S. growth is associated in the long run with an average 0.16 percent drop in growth across the sample. The findings based on a VAR approach for 46 countries, both advanced and developing, also suggest that U.S. growth disturbances have on average moderate dynamic effects on growth in other regions. The simulation results in Helbling and others (2007) also suggest that the potential spillovers from a temporary, U.S.- specific demand shock via trade channels alone are moderate, roughly of the same magnitude as the results from the panel and VAR analyses. However, alternative simulations, assuming correlated disturbances across countries, generate larger spillover effects. The authors of the study conjecture that such a higher impact of U.S. shocks could arise, for example, if the transmission had also involved asset price spillovers or confidence channels. Using a long-run (five-year average) panel regression approach for 101 countries over 1960–1999, Arora and Vamvakidis (2006) find much larger spillovers.

2 See, for example, Helbling and others (2007), Arora and Vamvakidis (2006), Bayoumi and Swiston (2009), and Swiston (2010). For an example of similar approaches applied to the case of China spillovers, see Arora and Vamvakidis (2010). 3 For the latter, see Bayoumi and Swiston (2009). This study examines the extent of spillovers across industrial regions including the U.S., the euro area, Japan, and an aggregate of small industrial countries, using VARs of growth across the four regions.

©International Monetary Fund. Not for Redistribution 102 Growth Spillover Dynamics: From Crisis to Recovery

In most specifications, a roughly 1 point increase in U.S. and EU growth is asso- 1 ciated with, respectively, a roughly 1 point and ⁄3 point increase in other countries’ growth (while Japan has an insignificant effect). However, the results could capture forces beyond short-term demand and business cycle effects transmitted via trade channels, such as technology spillovers, and are thus not directly comparable to those of Helbling and others (2007). Similar to Helbling and others (2007), Bayoumi and Swiston (2009) focus on short-run dynamics across industrial regions during 1970–2006. Their main results suggest significant U.S. spillovers to other industrial regions, including the eurozone, Japan, and a group of smaller advanced economies (of one-quarter to one-half the size of the U.S. shock after two years). Spillovers from the euro area or Japan are found to be smaller than spillovers from the United States and insignificant in most cases, except for spillovers from Japan to the euro area (which are found to be of similar size to those from the United States). Helbling and others (2007) generally find lower spillovers from the United States to other countries compared to Bayoumi and Swiston (2009). This is due to two main reasons: First, Helbling and others use the London Interbank Offered Rate (Libor) rate as a control variable. This amounts to some extent to shutting off a potential transmission channel and thus implies a lower estimated growth spillover from the U.S. GDP. In fact, Bayoumi and Swiston provide estimates of this transmission channel and show that the interest rate channel accounts for close to 50 percent of the transmission of U.S. growth shocks. Thus, once Helbling and others’ results are adjusted by this gap, spillover effects from the United States to other countries are closer to Bayoumi and Swiston’s estimates. Second, Helbling and others look at responses in countries which exclude the eurozone and Japan. As the authors themselves note, countries that have higher financial and trade linkages with the United States respond more strongly to growth shocks in the United States. In fact, for Canada and Mexico they find responses to U.S. growth shocks comparable in size to the average response found by Bayoumi and Swiston to U.S. shocks. Thus, the sample choice appears to explain the remaining difference in the estimated impact. Our results are comparable to Bayoumi and Swiston (2009) when looking at the entire sample period, but show an even higher response to U.S. GDP when looking at the more recent episode from 1993 to 2010. This largely reflects the higher financial integration of the United States with other countries in this period compared to the 1975–1993 period. To our knowledge, few studies have examined growth spillovers and linkages within Europe. For Germany, empirical results suggest relatively small growth spillover effects. Danninger (2008) estimates a VAR model for the growth rates of the United States, Japan, Germany, an aggregate of other euro area members, and an additional aggregate for the new EU member countries, for the period 1993–2007. While he finds significant spillovers from the euro area aggregate to the new EU member states and to Germany, he finds very limited spillovers from Germany to other countries. However, his estimation results for a more recent sample (since 1998) suggest stronger spillovers from Germany to the euro area

©International Monetary Fund. Not for Redistribution Poirson and Weber 103 and the new member countries, with 40 percent and 100 percent, respectively, of any German growth shock transmitted after three quarters. A related literature examines fiscal policy spillovers within Europe. Bénassy- Quéré and Cimadomo (2006) generally find positive cross-border spillovers from Germany, in the sense that a fiscal expansion in Germany raises GDP abroad, at least in neighboring and smaller countries. Similarly, Beetsma, Giuliodori, and Klaassen (2005) find that, averaged across all partner countries, the effect on foreign GDP of a fiscal stimulus in Germany of 1 percent of GDP is estimated to be 0.12 percent for a spending increase and 0.03 percent for a net tax cut. In Chapter 6 of this volume, Ivanova and Weber find spillovers of a similar order of magnitude for fiscal consolidation in Germany, France, the United Kingdom, and the United States. The authors argue that even when multipliers are very large, spillovers are limited, in particular from core EU countries to peripheral countries, since trade links between the two areas are not very strong. Few studies have examined the transmission channels of spillovers, that is, the major international channels through which shocks are propagated. Using model- based simulation analysis, Helbling and others (2007) find that most of the U.S. spillover effects are trade-related, and the effects are relatively small, roughly of the same magnitude as identified in the panel and VAR analyses. To generate larger effects, alternative simulations need to assume that disturbances are correlated around the world. The authors posit that such correlated disturbances could be related to increased trade and/or financial integration and could particularly arise in times of financial crisis. The simulation results in Bagliano and Morana (2011) also suggest a relatively more important role for the trade channel as a mechanism for transmitting U.S. economic developments to the rest of the world. Based on a large-scale open- economy factor VAR macroeconometric model, Bagliano and Morana find no clear-cut impact of adverse U.S. financial developments on foreign economic activity. While increases in a U.S. credit spread index lead to an output contraction abroad, U.S. stock price dynamics do not have any relevant effect on foreign GDP, and U.S. house price dynamics only affect the non-OECD group. Hence, the authors of the study conclude that the trade channel appears to be the key transmission mechanism of U.S. shocks to the rest of the world. By contrast, Bayoumi and Swiston (2009) find that the largest estimated contributions to spillovers come from financial rather than trade variables. Their result is based on a comparison of the response of GDP growth in a basic VAR model to that in a model augmented with a potential spillover source (either trade, commodity prices, or financial conditions) to measure the contribution of this source to the estimated spillovers. In particular, short-term interest rates and financial conditions more generally (bond yields and equity prices) are found to play an important role in the international transmission of U.S. growth shocks. Galesi and Sgherri (2009), using a global VAR approach, also report that equity prices are the main channel through which—in the short-run—financial shocks are transmitted from the United States to other countries. They find that other

©International Monetary Fund. Not for Redistribution 104 Growth Spillover Dynamics: From Crisis to Recovery

variables—including real credit growth, real GDP growth, and real interbank rate—become more important over a two-year horizon. The analysis of cross-country growth spillovers and, more generally, multicoun- try estimations is generally hampered by dimensionality constraints. Four different VAR-based approaches have been suggested to tackle this issue:4 Bayesian VARs, factor model VARs, global VARs, and VARs based on regional groupings. All four techniques also require an approach for resolving the identification issue. • The Bayesian VAR approach tackles the problem with the use of priors about the cross-country correlation patterns, which are subsequently updated with the data (Banbura et al., 2008, and Canova and Ciccarelli, 2006). • Factor models, instead, collapse cross-country co-movements of several variables into common factors that are then allowed to affect the dynamics of the individual countries (Bénassy-Quéré and Cimadomo, 2006). • Global VARs reduce the individual countries’ spillovers to their share in a weighted average for the variable of interest, which then affects the individual countries’ dynamics.5 The spillover in the global VAR therefore has a direct interpretation, unlike the spillover in the factor VAR (Bussière, Chudik, and Sestieri, 2009; Galesi and Sgherri, 2009; and Dees and others, 2007). • A fourth approach focuses on a small set of countries or regions—usually two to four—and then uses the traditional structural VAR (SVAR) approach (Bayoumi and Swiston, 2009; and Danninger, 2008). The degrees of freedom are preserved by reducing the number of regressors, that is, by reducing either the number of countries involved or the number of variables considered, or a combination of both. Bayesian VARs and SVARs are more general than global VARs or factor VARs, since they impose less structure on the inter-linkages. Compared to SVARs, Bayesian VARs require making more assumptions on the data generating process in return for more degrees of freedom, which makes the estimation feasible if the number of regressors is high relative to the size of the available data sample. The SVAR approach proposed by Bayoumi and Swiston (2009) requires an extensive dataset, but has the advantage that it imposes no structure on the inter-linkages, and thus the coefficient estimates are purely data-driven. Our analysis takes the existing literature further in the following way: • We confirm the finding of earlier studies, based on the assessment of spillovers across regions, that the United States remains the main source of growth spillovers, using an approach that involves a larger set of 17 individual countries.

4 Another possibility is to use model-based simulation analysis. See, for example, the analysis based on structural estimated macro models using panel unobserved components estimation as suggested by Vitek (2009 and 2010). 5 Cross-border trade weights are generally used to estimate the country-specific aggregate foreign variable, although one study uses annual bank lending exposures over 1999–2007 (Galesi and Sgherri, 2009).

©International Monetary Fund. Not for Redistribution Poirson and Weber 105

• We provide a decomposition of countries’ growth rate into the contributions from domestic and foreign components during the recent crisis and the recovery process. • We confirm the importance of financial transmissions channels for growth spillovers, but find that trade channels are also relevant for several countries. • A more general distinction made evident by the analysis in this paper is the distinction between autonomous country demand and transmission of global demand shocks. Our findings suggest that while spillovers vary positively with country size, they also reflect the extent to which growth is domestically driven. This has implications for the extent to which some economies can be considered engines of global or regional growth or, alternatively, can be considered transmitters of growth shocks that originate elsewhere.

EMPIRICAL APPROACH Given our focus on a univariate growth spillover framework, we follow Bayoumi and Swiston (2009) and minimize the structure we impose on the data. The resulting coefficient estimates thus capture all potential channels of transmission of shocks, including both trade and financial channels (the latter of which may be the most relevant in times of financial crises, when correlations between all risky assets tend to rise). The main specification is based on a reduced-form VAR estimation. Identification is obtained by weighting different orderings. The results from the different orderings can then be summarized by focusing on the average impulse response. This approach has the additional advantage that it provides not only a measure of uncertainty regarding the coefficient estimates but also a measure of uncertainty associated with the variation of responses across different orderings. The following derivations are reduced to a minimum when referring to the Bayoumi and Swiston (2009) approach.

Estimation Framework The general model is given by the following reduced-form model for the growth rate of output:

BLy()ttt=+ DLx () e where the vector y is given by stacking each country’s GDP growth rate ( yi,t): yy= y y tt()1, itIt , ,

We consider the following sample of 17 countries: Austria, Belgium, Canada, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, Portugal, Spain, Sweden, Switzerland, United Kingdom, and United States. The model is estimated on quarterly real PPP-adjusted GDP data, from 1975:Q1 to 2010:Q3, from the OECD Economic Outlook database. The control vector x includes two dummy variables for

©International Monetary Fund. Not for Redistribution 106 Growth Spillover Dynamics: From Crisis to Recovery

the oil shocks in 1979 and in 1990, respectively, and a constant term. When discussing the results we refer to the baseline regression as the regression which includes additionally a crisis dummy that takes the value 1 from 2008:Q4–2009:Q1, to reflect estimates of “normal times,” while the framework without the crisis dummy also reflects the relationship across growth rates in crisis times.6 Identification is obtained using the Choleski ordering of the countries in the sample, which provides the structural errors and coefficients: ε=−−11 =() = ()−1 () ttAeA(0) (l )AB (0) (lFl )AD 0 l

For the ordering, we distinguish three sets of countries according to their share in the total sample size, measured in USD PPP-adjusted GDP: Countries that surpass the 8 percent threshold are considered large countries and can be a leading country (U.S., Japan, and Germany). Countries that contribute less than 4 percent to the total output are considered small economies (Canada, Netherlands, Belgium, Sweden, Austria, Switzerland, Greece, Portugal, Ireland, and Finland). These latter countries are ordered last in the order of their size. Since we focus on shocks from the major countries to other nations, the ordering for the group of the remaining smaller countries does not affect the results. The intermediate group of countries comprises the medium-size countries (United Kingdom, France, Italy, and Spain). While they are never ordered first, they are generally ordered before the small countries. We arrange the orderings roughly according to the respective country’s relative size. The orderings assign a probability of 50 percent to the United States, being the lead country (in line with Bayoumi and Swiston, 2009), that is, the country that is not contemporaneously affected by other countries, and respectively a 25 percent probability to the United States to be ordered second or third. Japan and Germany are treated symmetrically throughout (although Japan is somewhat larger than Germany). Germany and Japan both have a chance of being ordered first, second, or third of 25 percent and a 12.5 percent probability of being ordered fourth or fifth. The United Kingdom, France, and Italy are also treated symmetrically, given their comparable average size. Each of these three countries 1 has a probability of being ordered second or third of 8 /3 percent, of being ordered fourth or sixth of 25 percent, and of being ordered fifth or seventh of 2 16 /3 percent. Spain has a probability of being ordered fifth or sixth of 25 percent and of being ordered seventh of 50 percent. This size-based procedure results in 48 different orderings, the details of which are provided in the Appendix.

6 Alternative control variables, including the oil and non-oil commodity price indices, U.S. and German short-term and long-term interest rates, U.S. investment grade and high-yield credit spreads, German corporate bond spreads, U.S. and German real equity prices, world trade, and Asia trade, were also included as a robustness check in an earlier version of the analysis. However, none of these control variables except the U.S. credit spreads and, to a lesser extent, U.S. real equity prices was significant, and their inclusion left the results unchanged. The impact of the U.S. credit spread, however, becomes insignificant when included in addition to the 2008–09 crisis dummy, suggesting that this variable is essentially a proxy for the global financial crisis.

©International Monetary Fund. Not for Redistribution Poirson and Weber 107

The computation of the standard errors and the ordering uncertainty follow Bayoumi and Swiston (2009).

Dynamic Growth Contribution The dynamic contribution is computed by applying to each ordering the following algorithm: Step 1: Determine the structural errors at each point in time using the Cholesky decomposition. ε=ˆ ˆ −1 ∀∈ ttAe(0) ˆ t T

Step 2: Compute the moving average (MA) representation of the entire history of each country’s growth rate.

T =εˆ yMTT∑ ()t −t t =0

Step 3: Create an identity matrix of dimension N and combine the MA representation and the structural errors to derive the contribution of the respective country’s shocks to the quarterly growth rate of GDP of the country under consideration.7

t ˆ =εˆ ˆ CMttt∑ ()t I − t =0

Step 4: Apply the relevant compounding rule to compute the annualized contribution.8 While the ordering of the small countries matters in theory for this exercise, in practice the ordering of these countries is of minor importance for the decomposition of the growth contribution, since growth dynamics are dominated by the larger economies. A similar calculation applies to the constant term, which allows decomposing the long-run growth rate into the contributions from the individual countries in the sample.

Transmission Channels To gain a better understanding of the transmission channels of the shocks, we employ a twofold strategy. The first approach uses a counterfactual analysis. We maintain the estimation framework as outlined above, but contrast the results to a scenario under which only the direct impact of the shock in country i is allowed

7 = ˆ Note that yCitt where i is a column vector of ones, with dimension N in the absence of any exogenous controls. 8 The rule will depend on whether the dependent variable and the shock are a level variable or a growth rate.

©International Monetary Fund. Not for Redistribution 108 Growth Spillover Dynamics: From Crisis to Recovery

to affect country j and all spillovers through third countries are prevented. This provides us with information about the relevance of inter-linkages across countries for the transmission of shocks. In a second approach, we estimate country- wise VARs, in which we include real exports as an explanatory variable to test for the importance of the trade channel. Counterfactual Analysis In the counterfactual analysis, we constrain the structural coefficients, which are associated with third-country effects, to equal zero. Thus, if we are interested in looking at the direct spillover from country i’s growth shocks to country k’s growth rate, we set all structural coefficients corresponding to the impact of country j’s growth shocks on country k equal to zero:9 α= kj ()l 0

for all l = 0,…,L and all j ≠ i and j ≠ k, where α are the structural coefficients. Impulse responses are then recalculated under the counterfactual assumptions. Country-by-Country Regressions To arrive at a parsimonious specification, we assume the sensitivity to specific countries to be homogenous within a region and the difference in the slopes within a group to be random. The United States, Japan, Canada, Sweden, Switzerland and the United Kingdom comprise the non-European Monetary Union (non-EMU, or RoW) group, and the rest of the countries constitute the EMU group. For each country we then run the following regression: = A()L yett

= RoWEMU EMU where the vector y is given by: yyt ()i,,t yi t EXPi ,,t yi t and Yi,t is EMU = k the growth rate of the output of the EMU area (YYi,tt∑ ) where k runs k≠i over all EMU members except for country i (when it is an EMU country) and Y RoW YYRoW = ∑ j i,t is the growth rate of the output of the non-EMU group ( i,tt) ji≠ where j runs over all non EMU members except for country i (when it is not an EMU country).10 While throughout this exercise we keep the ordering as above, we also allow in a robustness test for different orderings.11 This approach has the

9 This is likely to underestimate the relevance of the third-country effects since it is still possible that third countries can have a positive feedback to country i and via this to country k. In practice these are very small. 10 For all EMU members, the RoW shock is identical, while for all RoW members the EMU shock is identical. 11 In particular, we look at the average response to an EMU and a non-EMU shock ordering once RoW and once EMU first. This leaves the point estimates mostly unchanged. Allowing addi- Yi,t Yi,t = RoWEMU tionally for the following ordering Yt ()EXPi,,,t Yi t Yi t Yi ,t affects the point estimates, but not the relative magnitude across countries, leaving the interpretations unchanged.

©International Monetary Fund. Not for Redistribution Poirson and Weber 109 advantage that it provides a more convenient framework to distill the importance of the trade channel as a transmission channel and reduces the problem of ordering to a simpler choice. By grouping countries and using the group’s growth rate, we implicitly allow for an increasing weight of a country as it grows bigger relative to other countries. To see this, consider the coefficient on the growth rate of the non-EMU variable:

⎛⎞∑Y j t I ⎛⎞ ⎜⎟j≠k YY−− α=α−=α(0)y RoW (0) 1 (0)∑ i,1t ⎜⎟i ,1t − 1 kt, ⎜⎟j ⎝⎠ ⎜⎟∑∑YYt −−1,1i≠k j t Yi,1t − ⎝⎠j≠≠k j k

The right-hand version of the coefficient may be conveniently rewritten in the following form: I Y αα=αi,1t − ∑ i,,t(0)yi t i ,t (0) (0) i≠k ∑Y j,1t − j≠k

Thus, this model and the baseline estimation are identical under the assump- α=α=α = = tion that i,,t ()llli tt− ()i () for all lL1,.., and all t 1,..,T for this α =α + = =σ model and all ki ()llkk () v i with E()v 0 and E()vvv I in the baseline model and the extension that the real exports of country k are included as an additional variable.12

RESULTS This section provides an overview of the key transmission channels (trade vs. financial linkages) to help interpret the results, followed by a discussion of the main results derived from the estimation framework. The first set of results focuses on the potential impact of spillovers from a one percent growth shock in selected countries on other European countries. The second set of results presents the actual impact of selected countries on all other countries in the run-up to, during, and in the recovery from the recent financial crisis by combining the country-specific shocks with the impulse response functions. A third and fourth set of results, respectively, point to the relevance of the different transmission mechanisms of growth shocks and of different potential determinants of spillover size.

Cross-Border Linkages To interpret our results on the growth linkages, it is helpful to understand some key facts about the relative exposures of countries in the sample to the largest economies and regions. The most obvious channel is trade linkages: a rise in trading partners’ growth leads to an increase in their demand for imports, which

12 Unsurprisingly, this turns out not to hold, and the aggregation bias causes responses to be more pronounced (Imbs and others, 2005). Results are discussed in more detail in the respective section.

©International Monetary Fund. Not for Redistribution 110 Growth Spillover Dynamics: From Crisis to Recovery

TABLE 5.1 Exports to Selected Countries and Regions, 2010 (Percent of GDP) With With With With Euro United With United With With Rest of Germany Area1 States Kingdom Japan China World Austria 12.7 9.3 1.4 1.2 0.3 0.9 13.6 Belgium 16.5 37.0 4.6 6.2 0.8 1.4 20.1 Canada 0.2 0.8 18.4 1.0 0.6 0.8 2.8 Finland 2.9 6.4 2.0 1.4 0.5 1.4 14.0 France 3.2 6.5 1.0 1.3 0.3 0.5 6.8 Germany ... 15.7 2.0 2.4 0.4 1.8 14.2 Greece 0.8 2.2 0.3 0.4 0.0 0.1 3.2 Ireland 4.5 18.2 12.1 8.9 1.1 1.0 9.2 Italy 2.8 6.6 1.2 1.2 0.3 0.5 8.8 Japan 0.4 0.8 2.2 0.3 ... 2.7 7.8 Netherlands 18.8 26.2 2.7 5.6 0.5 0.9 17.7 Portugal 2.8 10.9 0.7 1.2 0.1 0.1 5.4 Spain 1.8 7.9 0.6 1.1 0.1 0.2 5.4 Sweden 3.5 9.9 2.1 2.6 0.4 1.0 13.5 Switzerland 7.2 10.9 3.8 2.2 1.2 1.4 10.6 United Kingdom 1.9 6.5 1.9 ... 0.2 0.4 5.8 United States 0.3 0.9 ... 0.3 0.4 0.6 6.1

Source: Direction of Trade Statistics (DOTS); IMF, World Economic Outlook; and IMF staff calculations. 1 Excluding Germany.

then contributes directly to an increase in the net exports of the home country (Table 5.1). With growing financial integration and cross-border ownership of assets, growth spillover effects may also be transmitted through financial linkages (Table 5.2). Both trade and financial exposures highlight the importance of intra- euro area transmission channels. For most European countries in our sample, the euro area as a whole is by far the biggest export market and accounts for the largest single banking sector exposure. The United States is also a key source of financial spillover risk for European countries. Within the euro area, it is noteworthy that for most countries (except the smaller trading partners), trade exposures to Germany are relatively smaller than trade exposures to the rest of the euro area as a whole, reflecting Germany’s relatively limited demand for imports from other European advanced countries. Several European countries, however, have large financial exposures to Germany. Trade exposures follow a strong regional pattern. In particular, they suggest a limited relevance of Asia for the sample countries (although growing in importance in the case of China). More specifically, we find that: • Trade links within Europe are important, as is evident in Table 5.1. The euro area is the largest export market for member countries in the sample, except Greece. For the Netherlands and Belgium, exports to the euro area account for about half of GDP, with a sizeable share (30 to 40 percent) directed to Germany alone.13 Even for Ireland, which has close trade ties with the

13 This holds also true when controlling for re-exports.

©International Monetary Fund. Not for Redistribution Poirson and Weber 111

TABLE 5.2 Banking Exposures to Selected Countries and Regions, 2010 (Percent of GDP)1 With With European Of which: With With Rest With Developed with Euro United With With United of Germany Countries2 Area3 States Japan China Kingdom World Austria 14.1 37.0 ... 4.7 0.4 0.5 5.4 67.1 Belgium 4.1 41.2 ... 8.1 0.2 0.3 8.5 22.0 Canada 1.0 4.3 ... 30.6 0.4 0.3 5.5 10.2 Finland 1.0 6.7 ... 0.2 0.0 0.0 1.0 1.3 France 11.0 54.0 49.4 21.5 6.0 0.8 12.4 23.7 Germany ... 41.2 37.8 15.7 1.7 0.5 14.4 16.6 Greece 1.8 7.1 ... 1.5 0.0 0.0 5.5 30.3 Ireland 21.7 67.2 ... 41.7 5.9 ... 91.0 31.5 Italy 12.9 15.2 14.1 2.1 ... 0.2 2.2 11.7 Japan 3.0 7.1 5.9 24.8 ... 0.7 2.9 5.4 Netherlands 22.0 57.6 ... 32.2 2.1 1.5 16.5 40.9 Portugal 2.0 31.7 ... 3.7 0.0 0.1 3.6 22.4 Spain 3.1 15.8 14.4 3.2 0.1 0.5 14.1 42.8 Sweden 16.3 94.3 ... 9.9 0.1 0.5 8.8 24.1 Switzerland 24.5 53.0 47.6 132.6 15.8 2.7 36.7 56.2 United 8.0 42.5 38.4 48.6 6.5 3.7 ... 62.2 Kingdom United States 1.6 5.1 4.1 ... 5.1 0.5 4.7 4.4

Source: Bank for International Settlements; IMF, World Economic Outlook; and IMF staff calculations. 1 International bank claims of domestically-owned banks, consolidated - ultimate risk basis. 2 Excluding Germany and the UK. 3 Excluding Germany.

United States, the share of exports going to the euro area exceeds that going to the United States. • The euro area is also the largest importer for the United Kingdom and Switzerland, and the second largest for the United States. However, it is only the fourth export destination for Japan, which relies more on U.S., Chinese, and other regional markets. • Despite its size and reflecting a relatively closed economy, the United States is generally not the major trading partner for European countries. • Trade exposures to Japan are even more limited. Trade exposures to China are also limited (except in the case of Japan), although the growing importance of China is underscored by the fact that China has now overtaken Japan in its importance as importer for all sample countries (except Ireland). • Similar to the United States and Japan, Germany’s relevance as an importer is relatively limited—including for other eurozone countries. While Germany is the second largest export destination for Austria and the Netherlands, it is only at best the third largest export exposure for the other European countries. These countries export relatively more either to the non-German euro area as a whole or to the rest of the world (excluding the

©International Monetary Fund. Not for Redistribution 112 Growth Spillover Dynamics: From Crisis to Recovery

euro area, the United States, the United Kingdom, China, and Japan). In the case of non-euro area countries, exports to Germany account for less than 0.5 percent of GDP. Financial linkages single out the United States and Europe, in the aggregate, as main sources of global spillover risks, while Germany’s importance is mostly regional. Overall, financial exposures—proxied by bilateral bank lending exposures on an ultimate risk basis—follow a broadly similar regional pattern as trade exposures, illustrating the relative importance of intra-European linkages: • For European countries, exposure to European developed countries is the single largest source of spillover risks, amounting to 30 percent of GDP or more (except in Finland and Greece).14 For the United States, exposure to Europe also ranks as the first source of risks; however, relative to GDP many European countries (including Switzerland, the United Kingdom, France, Germany, and the Netherlands) have much larger exposures to the United States than the United States does to Europe. • Within Europe, some countries have high exposures to Germany of 10 percent of GDP or more (France, Austria, Switzerland, Netherlands, Ireland, Sweden). By contrast, non-European countries (United States, Canada, and Japan) have banking exposures to Germany at or below 3 percent of GDP, well below their level of exposures to the United Kingdom in the case of the United States and Canada. • For Canada and Japan, financial linkages highlight the key role of the United States as a source of spillover risks, with European exposure the second largest source of risk. • Exposures to Japan are at or below 6 percent of GDP for all countries (except Switzerland), and exposures to China are below 1 percent of GDP in most cases (except the United Kingdom, the Netherlands, and Switzerland).

Country-Specific Outward and Inward Growth Spillovers We illustrate the potential cross-border spillovers by calculating the maximum weighted cumulative impulse response of the other countries to a one percent growth shock in the originating country. Detailed impulse responses of single countries are shown in the Appendix. Outward Growth Spillovers The baseline regression results (full sample, including a dummy for the 2008–09 crisis) suggest that the United States remains the largest source of spillovers to all countries in the sample. By contrast, Germany plays a minor role for the sample

14 In all cases, except for Greece, Spain, Austria, and the U.K., which also have significant exposures to the rest of the world (excluding the U.S., Japan, and Europe) and Switzerland (which has the largest single exposure to the U.S.), European developed countries are the single largest source of banking exposures.

©International Monetary Fund. Not for Redistribution Poirson and Weber 113 in aggregate, although some smaller European countries are strongly affected by German growth. The United Kingdom and Japan fall between the two, as do France, Italy, and Spain. Since the regressions include a crisis dummy, this first set of results can be interpreted as the “normal” response outside times of crisis. Regression results for an alternative specification, excluding the crisis dummy, show much higher outward spillover effects from all regions, supporting the view that nonstandard transmission channels are at play during times of crisis. Such channels could include, for example, correlated downturns in asset prices and/or confidence effects, which amplify the effect of a given growth shock over and above the standard transmission of shocks through trade and/or lending channels. While spillovers from all the large countries are found to increase during times of crisis, the increase is especially pronounced during the recent period for the United States, the United Kingdom, and the non-German euro area. This could reflect the fact that the 2008–09 crisis was triggered by the housing correction in the United States and the United Kingdom, which later affected other countries, like Spain, where the real estate market was similarly overheated. In contrast, Germany was not an independent source of shocks during the recent financial crisis. Estimates for the more recent subsample (1993:Q1–2010:Q3) do not suggest an increasingly important role of Germany for spillovers to other countries. By contrast, we find evidence of Italy’s and the United States’ increased importance for spillovers to other countries in recent years (Figure 5.2). Looking at disaggregated results within Europe, Germany’s role in generating outward spillovers appears limited despite the economy’s large size, in part reflecting Germany’s own dependence on growth in the rest of the eurozone (as discussed in the next section). Germany is particularly sensitive to growth shocks in the other three large euro area countries (France, Italy, and Spain) and in Japan, while the United Kingdom is less relevant. France is sensitive to Spain and, to a lesser extent, to Italy, and is much less sensitive to growth shocks in Germany than Germany is to growth shocks in France. Italy’s growth reacts relatively similarly to growth shocks in the other three large euro area countries, the United States, and Japan, but shows little sensitivity to the United Kingdom in normal times. Spain’s growth is potentially strongly affected by growth shocks in the United Kingdom and in France (possibly due to Spanish banks’ relatively large exposures to the United Kingdom) and to a lower extent by growth shocks in Italy. Germany and Japan generate only minor outward growth spillovers to Spain, reflecting limited trade and financial linkages. In particular, there is little support for the view that euro area peripheral countries could benefit strongly from Germany’s ongoing recovery. The results suggest that positive growth shocks in France and Italy generate larger spillover effects to the European periphery. The impact of Germany and other large countries on Greece, Ireland, and Portugal (GIP) can be summarized as follows: • Germany plays a less prominent role for the GIP than France and, to a lesser extent, than Italy. However, this result masks considerable heterogeneity in the responses across the GIP. Italy and France appear to have a relevant

©International Monetary Fund. Not for Redistribution 114 Growth Spillover Dynamics: From Crisis to Recovery

2.50 Response by all other countries to a growth shock in...

1993Q1:2010Q3 Excluding crisis dummy 2.00 1993Q1:2010Q3 1975Q1:2010Q3 Excluding crisis dummy 1975Q1:2010Q3

1.50

1.00

0.50

0.00 United States Japan Germany United Kingdom Rest of EAb GIP

2.50 Response of the GIP to a growth shock in...

1993Q1:2010Q3 Excluding crisis dummy 2.00 1993Q1:2010Q3 1975Q1:2010Q3 Excluding crisis dummy 1975Q1:2010Q3 1.50

1.00

0.50

0.00

United States Japan Germany United Kingdom France Italy

−0.50

Figure 5.2 Cumulative Peak Impulse Response after 10 Quarters to a 1 Percent Growth Shock (In percent)a Source: IMF staff estimates. aGDP-weighted average response of other countries. bExcluding Germany. influence on Ireland and Portugal. This is much less the case for Germany. However, Germany tends to have a higher impact on Greece than Italy does, but it still has less impact on Greece than France does. • The United States has much more of an impact on Ireland than on the other GIP countries, consistent with results in Kanda (2008) and the close trade and financial linkages between the two countries. The United States also has

©International Monetary Fund. Not for Redistribution Poirson and Weber 115

a more pronounced impact on Greece and Portugal in the context of the financial crisis and the more recent episode. • Japan and the United Kingdom play a relatively minor role for the GIP, although the United Kingdom’s impact during the crisis was more pronounced. Finally, the results suggest that potential negative growth spillovers from the GIP to other countries are limited, although a broadening of the crisis to larger eurozone countries would have significant real implications. The impact of Greece, Portugal, and Ireland as a group on other countries appears relatively small, consistent with their modest size. A shock to Spain’s growth, however, has potentially a much larger impact, particularly on other European countries (see Appendix). Inward Growth Spillovers Turning to the sensitivity of the four large euro area members to inward spillovers, the results underline Germany’s high sensitivity to external shocks. A first set of estimates from the small-country VARs shows that Germany responds to a growth shock in EMU countries more strongly than any of the other large EMU countries and exhibits the second largest response (after Italy) to growth shocks in non-EMU countries (Figure 5.3). Thus, Germany’s growth is more sensitive to foreign shocks than is that of other large EMU countries, especially to shocks originating within the EMU. This is consistent with Germany’s large trade and banking exposures to the rest of Europe. The result that Germany is highly sensitive to foreign shocks also supports the view that Germany is less of a source of

2 0.7 After 1 year After 10 quarters Large model 1.8 0.6 1.6

1.4 0.5

1.2 0.4 1 0.3 0.8

0.6 0.2

0.4 0.1 0.2

0 0.0 Germany Italy France Spain Germany Italy France Spain EMU shock Non-EMU shock

Figure 5.3 Sensitivity to a 1 Percent Growth Shock in the Large European Countriesa Source: IMF Staff estimates. a Right axis = weighted impact on growth in percentage points in response to shock under the large model; left axis = impact on growth in percentage points in response to shock under the small model.

©International Monetary Fund. Not for Redistribution 116 Growth Spillover Dynamics: From Crisis to Recovery

independent spillovers and more of a conduit for U.S. and other external shocks to the rest of Europe. A second set of estimates from the main baseline estimation (i.e., large model)15 yields the same rank ordering of sensitivity to external shocks as the smaller country-specific VARs.16 Regression results for the more recent period (since 1993, based on the large model) show that inward spillovers from the United States to the four large EMU member states have increased, while the United Kingdom and Japan have become less relevant (Figure 5.4). In the baseline estimation, in normal times a 1 percent growth shock in the United States tends to increase output growth within 10 quarters by about 0.3 percent in Germany, 0.4 percent in Italy and France, and 0.1 percent in Spain. These values increase to 0.4 percent in Germany and Spain, 0.5 percent in France, and 0.6 percent in Italy when not controlling for the effect of crisis times. The respective values more than double for Germany and Spain, in the more recent episode (estimation for 1993:Q1–2010:Q3 including a crisis dummy), as both become more sensitive to the United States than France and Italy. The increased sensitivity to the United States is in line with the increased correlation of the EMU countries’ GDP growth with the lagged GDP growth of the United States (Figure 5.1). This reflects rising financial linkages, which have become more important, particularly in the latter half of the sample period. It is consistent with similar findings by Helbling and others (2007), who find spillovers to be increasing in financial and trade linkages and significantly higher spillovers from the United States to other countries in the 1987–2006 period as compared to the entire 1970–2006 period. Since our sample includes years that mark the peak of financial linkages (2007 and 2008), we find even higher effects. Within the EMU, inward spillovers from Germany, Italy, and France to the other large EMU members are relatively stable across the two sample periods and are robust to the inclusion of the crisis dummy (i.e., likely to persist even outside of crisis times) (Figure 5.4). Specifically, we find that: • Germany’s effect on the other three large euro area countries ranges at the lower end for Spain and France, causing output there to rise by 0.1–0.2 percent, although Italy’s growth rate rises by around 0.4 percent in response to a 1 percent growth shock in Germany. • France affects Italy and Germany roughly by the same order of magnitude, yielding an increase in growth by 0.4 percent in normal times and above 0.5 percent in crisis times. The effect on Spain is only marginally higher. • A 1 percent growth shock in Italy causes German and French growth to increase by 0.4 percent and Spanish growth by 0.2 percent in the baseline

15 While estimates from the smaller model are directly obtained for “EMU” and “Non-EMU” shocks, the corresponding values from the baseline VAR (large model) is obtained by weighting the responses to the single countries’ shocks which constitute the EMU and the non-EMU group in the country- specific VARs. 16 However, it should be noted that the small country-specific VARs overestimate the impact, due to the aggregation bias which tends to increase the persistence of the shocks and thus overestimate the response.

©International Monetary Fund. Not for Redistribution Poirson and Weber 117

2.50 Inward Spillovers to Germany from Selected 2.50 Inward Spillovers to France from Selected Large Countries Large Countries

2.00 2.00 1993Q1:2010Q3 Excl. crisis dummy 1993Q1:2010Q3 Excl. crisis dummy 1993Q1:2010Q3 1993Q1:2010Q3 1975Q1:2010Q3 Excl. crisis dummy 1975Q1:2010Q3 Excl. crisis dummy 1.50 1975Q1:2010Q3 1.50 1975Q1:2010Q3

1.00 1.00

0.50 0.50

0.00 0.00 USA JPN DEU GBR FRA ITA ESP USA JPN DEU GBR FRA ITA ESP

−0.50 −0.50

2.50 Inward Spillovers to Italy from Selected 2.50 Inward Spillovers to Spain from Selected Large Large Countries Countries

1.00 1.00

0.50 0.50

0.00 0.00 USA JPN DEU GBR FRA ITA ESP USA JPN DEU GBR FRA ITA ESP

−0.50 −0.50

Figure 5.4. Inward Spillovers to the Four Large Euro Area Countries 1993–2010 and 1975–2010 Source: IMF Staff estimates.

estimation. The effect increases by an additional 0.2 percent for Germany and France and 0.3 percent for Spain in crisis times. In the more recent period, Spain’s sensitivity to Italy’s growth shocks has increased to a higher level than its sensitivity to growth shocks in Germany or France. • Spain has been an important source of growth shocks for other large euro area countries. Under the baseline regression for the full sample, a 1 percent shock to Spain’s growth increases GDP growth in Germany by 0.7 percent, in France by 0.5 percent, and in Italy by 0.3 percent. This position is reaf- firmed in the regression for the more recent episode, which shows a general increase in the importance of Spain’s growth spillovers. In particular, positive spillovers from Spain have become more important for France and Italy

©International Monetary Fund. Not for Redistribution 118 Growth Spillover Dynamics: From Crisis to Recovery

TABLE 5.3 Long-Run Growth Spillover (1975:Q1–2010:Q3) Percent spillover from: Average USA JPN DEU GBR FRA ITA ESP CAN Others growth USA 0.4 0.0 –0.2 –0.1 –0.1 0.3 0.1 0.0 3.0 JPN 0.5 –0.1 –0.1 0.0 –0.1 0.5 0.1 0.0 2.2 DEU 0.7 0.6 –0.1 0.1 –0.2 0.6 0.1 0.0 1.5 UK 1.5 0.3 0.0 0.1 –0.1 0.7 0.2 0.0 2.1 FRA 0.8 0.4 –0.1 –0.1 –0.1 0.5 0.2 0.0 1.9 ITA 0.9 0.6 –0.2 –0.2 0.1 0.4 0.2 0.0 1.5 ESP 0.7 0.3 0.0 –0.3 0.1 –0.1 0.2 0.1 2.1 CAN 2.1 0.3 0.0 –0.2 –0.2 –0.1 0.4 0.0 2.7 NLD 1.3 0.3 –0.2 –0.2 0.0 –0.1 0.6 0.2 1.9 BEL 0.7 0.4 –0.1 –0.1 0.1 –0.2 0.5 0.2 –0.1 1.6 SWE 1.4 0.2 0.1 –0.4 0.1 –0.2 0.6 0.2 0.0 2.1 Percent spillover to: spillover Percent AUT 0.5 0.2 –0.2 –0.1 0.2 –0.1 0.5 0.1 –0.5 1.3 CHE 0.8 0.4 0.0 –0.2 0.2 –0.3 0.5 0.1 –0.1 1.4 GRC 0.8 0.1 –0.3 –0.1 0.3 –0.1 0.8 0.1 –0.3 1.7 POR 0.2 0.5 –0.2 –0.1 0.4 –0.2 0.9 0.2 –0.7 1.7 FIN 1.6 0.4 0.2 0.3 0.3 –0.2 0.7 0.4 0.5 2.8 IRL 2.3 0.1 –0.1 0.2 0.2 –0.2 1.1 0.4 –0.7 3.6

Source: IMF staff estimates. Note: USA: United States; JPN: Japan; DEU: Germany; GBR: United Kingdom; FRA: France; ITA: Italy; ESP: Spain; CAN: Canada.

in recent years, consistent with the increase in trade and financial integration following EMU and euro adoption as well as with a domestic demand and property boom in Spain, both partly fuelled by unsustainable increases in corporate and household indebtedness.17

Domestic and Foreign Growth Contributions in Crisis and Recovery While the impulse responses measure the potential impact of growth spillovers across countries, they do not provide insight into the actual effect of idiosyncratic growth shocks originating in one country on other countries over time. The latter reflects both the size of the impulse response and the country-specific growth shock in the originating country. We investigate this question using a decomposition of the individual countries’ growth rates into the contribution from other countries over time. The results are summarized in Table 5.3, which provides estimates of the G7 individual countries plus Spain’s contribution to the other countries’ growth in the long run (i.e., in the absence of idiosyncratic shocks), and in the next series of graphs (Figure 5.5), which show the evolution

17 While the subsequent deflation of the property bubble and private sector deleveraging has resulted in negative dynamic contributions of Spain to other countries’ growth during the 2008–09 global recession, we find that on average, over the long run, Spain has been one of the major sources of positive growth spillovers to other countries, especially in Europe (see section D). However, the potential positive impact of Spain in future episodes could be lower than suggested by historical results if the unwinding of Spain’s imbalances is protracted and durably undermines Spain’s growth prospects.

©International Monetary Fund. Not for Redistribution Poirson and Weber 119

Germany France 8 8 6 6 4 4 2 2 0 0 −2 −2 −4 −4 −6 −6 −8 −8 −10 −10