51631 The World Bank AUGUST PREMnotes 2009 N U M B E R 141 ECONOMIC POLICY The Global Financial Crisis: Comparisons with the Great Depression and Scenarios for Recovery Milan Brahmbhatt (PRMVP) and Luiz Pereira Da Silva (DECVP) A recent paper by Eichengreen and O'Rourke 1. Comparing the Great on "A Tale of Two Depressions" (publicized Depression with the Present by Martin Wolf in the Financial Times) has Global Financial Crisis highlighted some close correspondences between economic performance during the Larger role of faster-growing developing present world recession and that during countries in the world economy the early months of the Great Depression Developing countries' share in world GDP that began in late 1929.1 World industrial was about 24 percent in 2008.2 Estimating production from April 2008 to April 2009 the share of the same countries in 1929 is fell as rapidly as during the first year of the not easy, but a rough calculation suggests Great Depression, while stock market prices around a 13 percent share at that time. and world trade volumes have fallen more China and India's combined share in world rapidly than in the comparable period. GDP was about 8.5 percent in 2008, versus These comparisons lead Eichengreen an estimated 2.8 percent in 19293 (figure and O'Rourke to draw the alarming conclu- 1). GDP share by itself would not necessar- sion that "[I]t's a Depression alright." They ily affect the evolution of the present global note, however, that fiscal and monetary poli- recession were it not for evidence that de- cies are likely to be much more supportive of veloping countries as a group are now also economic activity in the next 1­2 years than tending to grow substantially more rapidly they were during the first few years of the than developed countries. Great Depression. The first part of this note Figure 2 clarifies this statement. The outlines some other important structural dif- chart suggests that economic cycles in de- ferences between the world economy today veloping countries remain closely correlated and in the 1930s that are likely to affect how with those in developed countries. Devel- the present recession plays out relative to the oping country growth has fallen sharply Great Depression. The second part of the in 2009 through a variety of well-known note discusses possible recovery paths out channels, such as declining exports to devel- of the current crisis. oped countries, precipitous falls in private FROM THE POVERTY REDUCTION AND ECONOMIC MANAGEMENT NETWORK Figure 1. China and India : Share of World GDP (%) (Constant 2000 Market Prices and Exchange Rates) 9 8 7 6 Percent 5 4 3 2 1 0 1929 1938 1950 1960 1970 1980 1990 2000 2008 Year Source: Estimates for 1929 through 1950 are by World Bank staff, drawing on data in Angus Maddison, 2003, The World Economy, Historical Statistics, OECD, Paris. capital flows to developing countries, and such it should persist in the medium term, weakening remittance flows. In other words despite the severe negative shock of the pres- there has been no decoupling in the cyclical ent crisis.4 Another hypothesis is that the component of developing country growth. growth improvement of recent years mainly However figure 2 also suggests that, while reflected the extraordinary but temporary before the early 2000s the trend rate of boom or "bubble" conditions in the world growth in developing countries was close to economy, including very low international that in developed countries, since then trend interest rates and huge but unsustainable growth in developing countries has become credit flows to developing countries. In this substantially higher than in the advanced case we should expect the recent higher world. In other words, there has arguably growth of developing countries to quickly been a decoupling in underlying trend rates disappear, along with the bursting of the of growth. Thus, for example, developing bubble and the present global recession. country growth averaged only 0.8 percent- It is doubtless too early to fully judge age points higher than developed countries which if either of these hypotheses is correct. in the 1990s, but this growth gap widened to There are, nevertheless, some significant 3.5 percentage points in 2000­08. pieces of evidence in favor of the first hy- One hypothesis is that the rise in de- pothesis. All major forecasters expect devel- veloping country trend growth over the last oping countries to continue growing much decade is mainly a payoff for the strenuous faster than developed countries even during efforts by many of these countries to improve the depths of the present crisis. For example, their macroeconomic, structural, and other World Bank projections are for developed policies over the last 2­3 decades, accom- country real GDP to shrink by 4.2 percent panied over the past decade with a marked in 2009 while developing country growth is improvement in country balance sheets. As expected to hold up at a positive 1.2 percent, 2 PREMNOTE AUGUST 2009 Figure 2. World Output Growth 1961 -2011 (% Change) 10 Developed countries 8 Developing countries 6 4 Percent 2 0 ­2 ­4 ­6 61 64 67 70 73 76 79 82 85 88 91 94 97 00 03 06 09 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 20 Year Source: World Bank WDI and GDF. a 5.4 percentage point positive growth gap. greater importance of the services sector in If these projections are roughly correct, the major advanced economies today. The then the positive growth gap for developing simplest way to draw this distinction is to countries and their larger share in world ac- look at the share of services in employment tivity together would have helped to reduce (figure 3). In the United States services the size of the contraction in overall world rose from 55 percent of total employment output in 2009 by a significant amount--by in 1929 to 82 percent in 2007 (of which 66 about 1.3 percentage points--from a po- percent is in private sector services and 16 tential 4.2 percent fall to the smaller 2.9 percent in government employment). percent global contraction that is currently Services are also considerably less vola- projected. Recent data indeed suggest that tile than goods sectors (figure 4). Employ- growth in at least some emerging economies ment volatility in the U.S. private services has rebounded with unexpected strength sector averaged only half that in goods pro- in the second quarter of 2009, averaging duction over 1960­2007, while volatility around 10 percent at annualized quarter on of government employment was only 40 quarter rates among several emerging East percent as high.5 This provides a reason Asian economies, for example. to conjecture that aggregate employment volatility today might be a good deal lower Larger share of less volatile than in the 1930s. services in global activity Another angle on the lower volatility of There are other differences in the structure services is to look at the behavior of real out- of the global economy between now and put in the goods and services sectors during 1929 that tend to limit the usefulness of recessions. Table 1 shows the percent change industrial production alone as an index for in real output in the four quarters after the overall economic activity, in particular the start of a recession, for the last six U.S. reces- AUGUST 2009 PREMNOTE 3 Figure 3. United States--Sector Employment Shares, 1929­2007 (% of total) 70 60 50 40 Percent 30 20 10 Goods Private services Government 0 19 9 19 1 19 3 19 5 19 7 19 9 19 1 19 3 19 5 19 7 19 9 19 1 19 3 19 5 19 7 19 9 61 19 3 19 5 19 7 19 9 19 1 19 3 19 5 19 7 19 9 19 1 19 3 19 5 19 7 19 9 19 1 19 3 19 5 19 7 20 9 20 1 20 3 20 5 07 2 3 3 3 3 3 4 4 4 4 4 5 5 5 5 5 6 6 6 6 7 7 7 7 7 8 8 8 8 8 9 9 9 9 9 0 0 0 19 19 Year Source: U.S. Bureau of Economic Analysis. Tables 6.5A­D. World Bank staff calculations. sions. The table shows that all of the output we only consider three quarters through contraction in U.S. recessions occurs in the the first quarter of 2009. Nevertheless, one goods and structures sector. The services might hope that the greater weight of the sector shows positive growth throughout. more stable services sector will moderate The table shows how the present recession overall output volatility compared to the is already more severe than the previous five Great Depression. studied in Table 1, even though in this case Figure 4. United States--Employment Volatility, 1960­2007 0.060 Goods 0.050 Private Services Government Standard deviation 0.040 0.030 0.020 0.010 0.000 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 00 02 04 06 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 20 Year Source: U.S. Bureau of Economic Analysis. Tables 6.5A­D. World Bank staff calculations. 4 PREMNOTE AUGUST 2009 Table 1. Percent Change in Output Four Quarters after Preceding Peak in Six U.S. Recessions Date of preceding peak 1973:4 1980:1 1981:3 1990:3 2000:4 2008:2 * Goods and structures -7.0 0.9 -6.9 -2.7 0.8 -8.7 Services 3.0 1.8 1.7 1.8 2.6 0.8 Source: U.S. Bureau of Economic Analysis, Table 1.2.6. * Change in 3 quarters to the first quarter of 2009. Changes in the structure of world trade Policy responses: then and now Another structural difference between "now As noted, Martin Wolf and Eichengreen and and then" worth probing relates to world O'Rourke both highlight the much stron- trade volume, which, as Eichengreen and ger counter-cyclical macroeconomic policy O'Rourke observe, has fallen more in the response of the present day as a major posi- present recession than during the first year tive difference with the Great Depression. of the Great Depression. However, recent Here we only supplement their discussion World Bank research suggests that the with some additional observations on policy sharper recent trade decline compared to responses. 1929/30 may to some extent reflect a much A key difference between the 1930s and greater responsiveness or elasticity of trade the current crisis is the role of the gold stan- with respect to GDP, rather than only a dard as a deflationary force. In the 1930s, steeper decline in GDP driving a larger fall most economists and policy makers saw the in trade.6 This research points to a sharp gold standard as the guarantor of economic rise in the elasticity of world trade to GDP stability. However, adherence to the gold from under 2 in the 1960s to over 3.5 now, standard proved to be quite dysfunctional with even an even sharper responsiveness under the economic conditions of the early during downturns than during tranquil 1930s. The U.S. Federal Reserve, for exam- times. The reasons for this large increase in ple, dramatically raised its rediscount rate in trade elasticity are not altogether clear, but 1931 in response to Britain's devaluation (in one explanation put forward in the cited conformity with standard classical theory), research is the rise of modern fragmented fostering a significant reduction in the sup- production processes, which result in vastly ply of money and further compromising the increased cross-border flows of intermediate health of the banking sector. This, in turn, inputs used in the production of any final helped to transform a significant recession product. The importance of such processes into the Great Depression. in the world economy and world trade has The danger of repeating the trajectory increased dramatically in recent decades. of the 1930s because of dysfunctional mon- Today, declines in trade are likely to reflect etary policy reactions is less today, with the much smaller declines in production value prevalence of flexible exchange rate regimes added than was in the case in the 1960s and, between major economies (such as between one could argue, a fortiori, even more than the United States, the euro zone, and Japan), in 1929. The brighter side of the coin is that and also for a significant number of devel- we should also see a much sharper upswing oping economies. Policy makers now have in world trade when the recovery in world greater autonomy to target monetary policy output begins. at domestic activity and inflation/deflation AUGUST 2009 PREMNOTE 5 Figure 5. Central Banks' Total Assets (12/29/06 = 100) 350 United Kingdom 300 United States 250 12/29/06 = 100 Euro area 200 Collapse of Lehman Brothers 150 100 Japan 50 1/ /06 /3 7 31 7 /3 8 31 8 /3 7 /3 8 2/ /07 3/ /07 4/ /07 5/ /07 6/ /07 7/ /07 8/ /07 9/ /07 /3 7 2/ /08 3/ /08 4/ /08 5/ /08 6/ /08 7/ /08 8/ /08 9/ /08 /3 8 2/ /09 3/ /09 9 11 1/0 1/ 1/0 11 1/0 1/ 1/0 12 0/0 12 0/0 10 0/0 10 0/0 /0 9 29 28 31 30 31 30 31 31 29 31 30 31 30 31 31 28 31 /2 3 3 12 Year Source: IMF World Economic Outlook, April 2009. concerns. The rapid and comprehensive recognize that policy responses have been intervention by monetary authorities in effective in avoiding a financial collapse and the present recession is reflected in the in at least providing some underpinning for dramatic expansion of central bank balance global demand, although there remain vig- sheets shown in figure 5. Policy makers have orous debates over how effective monetary used almost every conceivable policy tool and fiscal stimulus will ultimately prove. to provide support for troubled financial In addition, there are significant concerns sectors: capital injections, direct purchase about how and when to design "exit policies" of troubled assets and discretionary lend- from the current stimulative monetary and ing by the Treasury, central bank support fiscal policy stance and about the potential with Treasury backing, liquidity provision, for longer-term weakening of fiscal positions guarantees, upfront government financing, as a result of current policies. and sometimes outright nationalization and liquidation of financial institutions. 2. Scenarios for Recovery The volume of such support has been un- from the 2008­09 Global precedented, reaching about 50 percent of Financial Crisis advanced economy GDP.7 The potential pace and path of a recovery In addition to support for the financial from the present global recession are now sector, policy makers have generally ac- beginning to emerge as important questions. cepted--though some reluctantly--the need The Great Depression again provides a use- for significant fiscal stimulus policy action. ful point of reference. In the United States, In most cases, G20 countries have adopted for example, real GDP fell 8.6 percent in discretionary fiscal stimulus measures that 1930 and continued to fall through 1933, reached 0.5 percent of their average GDP four years in which output fell a cumulative in 2008, 2 percent in 2009, and a projected 26.5 percent. The economy bottomed dur- 1.5 percent in 2010 (figure 6). Most analysts ing the course of 1933 with recovery setting 6 PREMNOTE AUGUST 2009 Figure 6. G20--Fiscal Stimulus and Financial Sector Support Turkey Saudi Arabia Russian Federation Indonesia Headline support to the financial sector (% of 2008 GDP) a India China Estimated cost of fiscal discretionary measures (% of GDP, relative to 2007 baseline) Brazil Argentina Korea, Rep. of Japan Australia United Kingdom Italy Germany France United States Canada 0 10 20 30 40 50 60 70 80 90 Percent of GDP Source: IMF Staff Position Note SPN/09/13, June 9, 2009. a. Capital injections, purchase of assets and lending by Treasuries, central bank liquidity, and other support and guarantees excluding deposit insurance. in from 1934 and continuing through 1937, protracted as during the Great Depression. followed by renewed but relatively brief re- Nevertheless studies have noted that the cession in 1938 (figure 7). characteristics of the present recession--for Few expect the path of recovery from example that it combines a credit crunch the present recession to be as delayed or and financial crisis with busts in the hous- Figure 7. U.S. Real GDP 1929­1939 (billion 2000 dollars) 1,000 900 800 US$ billion 700 600 500 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 Year Source: U.S. Bureau of Economic Analysis. AUGUST 2009 PREMNOTE 7 Figure 8. Stylized Scenarios for the Recovery Output Fast Recovery (1) Past Trend Growth Post-Crisis New Growth Path (2) Stagnation (3) "Green Shoots" Double Dip Recession (4) Time ing and equity markets and that it is highly would be proven wrong and that, despite the synchronized across countries--are such as damage to private financial wealth and to to suggest that this recession will be deeper bank and corporate balance sheets, growth and more protracted than any in the post- in private demand and financial sector lend- war period.8 However, with high continued ing is able to rebound along the lines of past uncertainties on many fronts, one can at this "normal" business cycles. The driving force stage only lay out some stylized scenarios behind such a scenario would need to be a for potential paths out of the recession, much faster than expected healing of the identifying some of the key variables to financial sector and its capacity to resume observe in the months to come that might lending at more normal levels and condi- help identify which (if any) of these scenarios tions (for example through a much more is coming to pass. Figure 8 sketches four radical restructuring and recapitalization possible scenarios for the coming one to of insolvent institutions using public funds two years (and perhaps beyond), with the than appears to have been politically feasible present (July­August 2009) highlighted as in many cases). Consumers and investors the point where many commentators have would also likely need to have "very short noted the emergence of some "green shoots" memories," taking a much more confident of incipient recovery.9 view of the future than would appear war- ranted by the dramatic events of 2008, Fast or "V-shaped" recovery continued uncertainties, and their own de- A fast or "V-shaped" recovery would mean pleted net worth positions. There could be that, after all, most analysts of this crisis specific countries that follow this recovery 8 PREMNOTE AUGUST 2009 pattern, but, given what we are observing in spring and summer rally in the global equity private consumption, investment, and credit and debt markets; an inability to implement markets, coupled with a proposed tighten- sufficiently strong fiscal stimulus programs, ing of prudential and regulatory standards, perhaps due to political factionalism and it seems that this type of a recovery is less deadlock in some countries; and a strong likely in general. revival of protectionism. Stagnation scenario New post-crisis moderate A stagnation scenario would be a replica of growth scenario Japan's Great Recession of the 1990s at a With all due caveats, a new post-crisis mod- global level. The driving force behind this erate growth scenario is perhaps the most scenario would be that deleveraging in the likely, with a recovery beginning in the financial sector continues, credit markets latter part of 2009 but growth during the remain stalled, and economic activity can- recovery failing to reach previous recovery not resume at previous levels, given low trends. It is also the scenario underlying confidence among consumers and inves- most mainstream projections, such as those tors, despite recent policy stimulus efforts. of the IMF, World Bank, and OECD. The Significant fiscal stimulus efforts tend to be main driving forces behind this scenario in- ineffective because of offsetting behavior clude a substantial rise in the propensity for by consumers and firms. As in Japan in the private savings as households attempt to pay 1990s, debt repayments put a downward off debt and build up their net worth, and, pressure on economic activity (irrespective of concomitantly, a need to rebalance growth low borrowing costs) and encourage higher in the global economy and reduce excessive private savings and a lower appetite for new current account imbalances. This correction investment.10 will take many forms (and some time), in- cluding changes in private consumption in "Double dip" recession scenario the United States but also in some surplus A "double dip" recession scenario would rep- countries, movements of relative prices resent the closest proxy today of the Great (for energy, minerals, food, and so forth), a Depression. It should be recalled that in the higher cost of capital due to deleveraging, 1930s there were several episodes of stock and tighter control of the financial sector. market recoveries between 1930 and 1932 These changes will all contribute to lower ("bear market" rallies) that created some trend growth for the next cycle. illusion of a faster recovery without eventu- Another reason this scenario is likely ally triggering a real economy response.11 is the need, at some point, for the current In addition to the forces operating in the monetary and fiscal policy stimulus to be "stagnation" scenario, the driving forces unwound. That will take out some momen- behind this possible replica of the Great tum from global growth and will have to be Depression could be the combination of a carefully timed to avoid both the dangers of new wave of financial trouble, perhaps in the inflationary pressure and premature weaken- euro area following banking and currency ing of aggregate demand. New investment crises in Eastern Europe, or as a result of will also be hampered by the large amount a large and disorderly U.S. dollar devalu- of excess capacity in the world economy and ation. For example, we could see renewed the time that will be required to run this risk aversion and a jarring end of the recent capacity down. Lastly, very severe recessions AUGUST 2009 PREMNOTE 9 tend to reduce--at least temporarily--trend 4. The case for a positive `trend decoupling' in growth at the beginning of the new cycle. It developing countries' growth is argued in detail might take some time before Schumpeterian by Jonathan Anderson, 2009, The Real Decoupling, "creative destruction" kicks back in to raise UBS Investment Research, Emerging Economic Perspectives Number 7 (August). total factor productivity, especially if the 5. Volatility measured as standard deviation recession takes a toll on R&D expenditure at (over a five-year rolling window) of the cyclical the firm level (even if public programs can component from a Hodrick-Prescott filter, as a compensate for part of the losses). ratio of the Hodrick-Prescott trend. There remain many challenges for poli- 6. Caroline Freund, 2009. "The Trade Re- cy makers and international financial institu- sponse to Global Downturns: Historical Evi- tions in this transition period from "green dence," World Bank DEC Research Group (June), shoots" to the next phase. One of the main World Bank, Washington, DC. post-crisis considerations for developing 7. See IMF, 2009, "Fiscal Implications of the countries will be the new productivity trend Global Economic and Financial Crisis," IMF of their economies in the wake of this crisis. Staff Position Note SPN/09/13 (June 9), IMF, Those who have targeted fiscal stimulus Washington, DC. well, reduced (infrastructure) bottlenecks, 8. Stijn Claessens, M. Ayhan Kose, and Marco Terrones, 2008, "What Happens During Reces- and invested wisely in human and physical sions, Crunches and Busts?" IMF Working Paper capital without exceeding their fiscal room 08/278. Carmen Reinhart and Kenneth Rogoff, for maneuver will be naturally better off and 2008, "The Aftermath of Financial Crises," CEPR ready to start a more robust growth cycle, Discussion Paper 7209, Center for Economic using their (new) comparative advantages Policy Research, London, UK. in the global economy. 9. See a similar description by Samuel Brittan, 2009, Financial Times (June 24). Notes 10. R. C. Koo, 2008, The Holy Grail of Mac- 1. Martin Wolf, 2009, "The Recession Tracks roeconomics, Lessons from Japan's Great Recession, the Great Depression," Financial Times, June 16, John Wiley & Sons. Koo stresses that the collapse citing Barry Eichengreen and Kevin O'Rourke, of the Japanese bubble did not result in a con- 2009, "A Tale of Two Depressions," June 4, www. traction of Japanese output after 1990 thanks to voxeu.org. massive fiscal spending. He adds that the crisis 2. Measured in real terms, at 2000 prices and triggered a behavioral change in private firms market exchange rates. The developing country and consumers (for example, deleveraging), af- share in 2008 in nominal terms at market ex- fecting the transmission and the role of monetary change rates was a little higher, about 27 percent. policy in Japan. Source: World Bank World Development Indica- 11. See Simon Hunt Strategic Services, 2009, tors and staff estimates. "April Economic Report," Weybridge, Surrey, 3. Estimates for 1929 through 1950 are by England. World Bank staff, drawing on data in Angus Maddison, 2003, The World Economy, Historical Statistics, OECD, Paris. This note series is intended to summarize good practices and key policy findings on PREM-related topics. The views expressed in the notes are those of the authors and do not necessarily reflect those of the World Bank. PREMnotes are widely distributed to Bank staff and are also available on the PREM Web site (http://www.worldbank.org/prem). If you are interested in writing a PREMnote, email your idea to Madjiguene Seck at mseck@worldbank.org. For additional copies of this PREMnote please contact the PREM Advisory Service at x87736. PREMnotes are edited and laid out by Grammarians, Inc. Prepared for World Bank staff