Financial Crises
Chapter

Chapter 9. What’s the Damage? Medium-Term Output Dynamics after Financial Crises

Author(s):
Stijn Claessens, Ayhan Kose, Luc Laeven, and Fabian Valencia
Published Date:
February 2014
Share
  • ShareShare
Show Summary Details
Author(s)
Abdul Abiad, Ravi Balakrishnan, Petya Koeva Brooks, Daniel Leigh and Irina Tytell The authors would like to especially thank David Romer for his many insights and suggestions. They are grateful also to Olivier Blanchard, Charles Collyns, Jorg Decressin, and to participants at an IMF Research Department seminar. Chris Papageorgiou kindly provided the authors with computer code to implement the Bayesian model averaging analysis. Stephanie Denis, Murad Omoev, and Min Song provided excellent research assistance. This chapter expands on ideas presented in Chapter 4 of the October 2009 World Economic Outlook (IMF, 2009b).

The recovery from the global financial crisis has been slow and bumpy, with unemployment remaining at high levels, and there are concerns about the prospect of long-term damage to economic activity. In this context, the aftermath of past financial crises may provide useful insights into the medium-term prospects for economies that recently experienced a crisis. To shed light on the medium-term outlook for such economies, this chapter examines the aftermath of banking crises in advanced, emerging, and developing economies over the past 40 years.

A first glance at several previous episodes suggests that although banking crises typically lead to large output losses in the short term, what happens to output in the medium term has varied widely (Figure 9.1). Some countries persistently grow at a slower rate than before, moving farther away from their precrisis trends, as in Japan and Thailand in the late 1990s. Some return to the precrisis growth rate, but fail to recover the initial output loss, as in Sweden (1991) and the Republic of Korea (1997). Some eventually return to their precrisis trend (Turkey, 2000), and some recover quickly and outperform the previous trend (Mexico, 1994).

Figure 9.1Selected Banking Crises

(Log scale)
(Log scale)

Sources: World Bank, World Development Indicators;Laeven and Valencia (2008) for banking crisis dates; and IMF staff calculations.

Note: See text for definition of precrisis trend.

A great deal of work exists on the output effects of financial crises in the short term, but until recently, the emphasis on the medium term following banking crises has been much more limited, with the notable exceptions of Boyd, Kwak, and Smith (2005), Cerra and Saxena (2008), and Reinhart and Rogoff (2009a).1 Given the banking crises in a number of economies, including the United States, in the aftermath of the 2007–09 global economic and financial crisis, interest in the topic has risen. For instance, Furceri and Mourougane (2009) apply the Cerra-Saxena approach, which involves using an autoregressive model of output growth rates augmented by crisis dummies, to growth rates of potential output for Organization for Economic Cooperation and Development countries. Pisani-Ferry and van Pottelsberghe (2009) also discuss the persistent impact on output of banking crises using several case studies. In another study, Haugh, Ollivaud, and Turner (2009) analyze the impact of banking crises on potential growth in Finland, Japan, Norway, and Sweden.

This chapter extends those studies in six main ways. First, it examines the medium-term dynamics of output in a particularly broad sample that includes 88 banking crises over the past four decades and across countries with high, middle, and low income levels. Second, it explores not only how the postcrisis level of output compares to the precrisis trend (output loss), but also how the postcrisis growth rate of output compares with its precrisis trend growth rate (growth loss). Third, with regard to methodology, the estimation of the precrisis trend ends several years before the crisis so that it is not contaminated by the possibility of an unsustainable boom in the run-up to the crisis or a precrisis slowdown. Fourth, the analysis decomposes the medium-term dynamics of output into both factor components (capital, employment, labor force participation, and total factor productivity) and demand-side factors (consumption, investment, exports, and imports). Fifth, given the wide range of postcrisis outcomes, the analysis assesses the correlation between postcrisis output and growth losses and variables measuring initial conditions and policy responses. Finally, with five years of data since the start of the global financial crisis, some initial evidence is provided on the medium-term implications of the most recent banking crisis episodes.

The first main finding is that the path of output tends to be depressed substantially and persistently following banking crises, with no rebound on average to the precrisis trend in the medium term. Seven years after a crisis, the level of output has typically declined by about 10 percent relative to the precrisis trend. Growth does, however, tend to eventually return to its precrisis rate.

Second, the depressed path of output tends to result from long-lasting reductions of roughly equal proportions in the employment rate, the capital-to-labor ratio, and total factor productivity. In the short term, the output loss is mainly accounted for by total factor productivity, but unlike the employment rate and capital-to-labor ratio, the level of total factor productivity recovers somewhat to its precrisis trend in the medium term. In contrast, capital and employment suffer enduring losses relative to trend.

Third, initial conditions appear to have a strong association with the size of the output loss. What happens to short-term output is also a good predictor of the medium-term outcome, as is the joint occurrence of a currency crisis and a banking crisis. This is consistent with the notion that the output drop is especially persistent following large shocks, carrying over into the medium term. A high precrisis investment share is a reliable predictor of high medium-term output losses, through its correlation with the dynamics of capital after the crisis. Evidence also suggests that limited precrisis policy space tends to be associated with more muted medium-term recoveries. An interesting finding is that postcrisis output losses are not significantly correlated with the level of income.

Fourth, the medium-term output loss is not inevitable. Some countries succeed in avoiding it, ultimately exceeding the precrisis trajectory. Although post-crisis output dynamics are hard to predict, the evidence suggests that economies that apply countercyclical fiscal and monetary stimulus in the short term after the crisis tend to have smaller output losses in the medium term. There is also some mixed evidence that structural reform efforts are associated with better medium-term outcomes. In addition, a favorable external environment is generally associated with smaller medium-term output losses.

Finally, the performance of economies that experienced banking crises during 2007–09 bears a sobering resemblance to performance in previous banking crises. These economies have experienced even deeper output losses, averaging 17 percent relative to the precrisis trend as of 2012. Moreover, little evidence indicates that output is returning to the precrisis trend.

How do these findings relate to shifts in potential output following financial crises? The term “potential output” typically refers to the level of output consistent with stable inflation and is a function of structural and institutional factors. A medium-term decline in output relative to the previous trend could reflect a decline in potential output, but it could also partly reflect a persistent fall in aggregate demand. The experiences of a number of economies, including Japan, suggest that if output remains below its precrisis trend in the medium term, a substantial part of the shortfall reflects lower potential. Therefore, to the extent that this chapter identifies output losses seven years after a financial crisis, it is likely that lower potential explains a substantial part of those losses. However, attempting to identify precise shifts in potential output is beyond the scope of this chapter.

The chapter is organized as follows. The first section describes key features of medium-term output dynamics following banking crises based on international experience during the past 40 years. The second section decomposes medium-term output losses into their factor components (capital, labor, and productivity), as well as their demand-side drivers (consumption, investment, exports, and imports). The third section analyzes the way in which medium-term output losses relate to country characteristics and macroeconomic conditions prevailing before the crisis. It also examines the role of domestic policies and the external environment after the onset of the crisis. The fourth section compares the experience through late 2012 of economies that entered banking crises during 2007–09 with the historical pattern. The last section concludes the chapter.

Does Output Recover in The Medium Term?

This section presents key stylized facts on the output losses associated with banking crises. It starts with methodological issues and then reports stylized facts on the estimated output losses at both the country level and the global level.

The analysis focuses on banking crises, and uses a comprehensive set of banking crisis events from Laeven and Valencia (2008) from the early 1970s through 2002. The Laeven-Valencia data set is constructed by combining quantitative indicators measuring banking sector distress, such as a sharp increase in nonperforming loans and bank runs, with a subjective assessment of the situation. Currency crises are also considered for purposes of comparison; currency crisis dates are identified based on the methodology of Milesi-Ferretti and Razin (1998). This definition requires (1) a 15 percent minimum rate of nominal depreciation with respect to the U.S. dollar; (2) a minimum 10 percent increase in the rate of depreciation with respect to the previous year; and (3) a rate of depreciation of less than 10 percent in the previous year. The sample includes 88 banking crises and 222 currency crises, distributed across countries with high, middle, and low incomes. The sample excludes transition countries because the output developments observed in these economies were strongly related to the shift away from central planning rather than to financial crises.2 Countries with populations of less than 1 million are also dropped.

The medium-term output loss for each episode is computed as illustrated in Figure 9.2. The idea behind the exercise is to measure the output loss associated with a crisis as the difference between the actual level of output and the level that would have been expected based on the prevailing precrisis trend. In line with the focus on the medium term, the analysis uses a postcrisis window of seven years, looking beyond the effects of short-term fluctuations of the economy. In addition, because it is possible that the slope of the trend may be affected by the crisis, growth losses are computed as the difference between the growth rate after the crisis and the precrisis trend growth rate. The precrisis trend growth rate is defined as the slope of the precrisis trend depicted in Figure 9.2.

Figure 9.2Output Loss Methodology Example

(Republic of Korea 1997)

Sources: World Bank, World Development Indicators; and IMF staff calculations.

Note: First year of crises at t = 0. The precrisis trend is estimated up to year t = –3, and is extrapolated linearly thereafter. Trend equals 100 in year t = 7.

Estimating the precrisis trend is challenging because the analysis needs to be insulated from the impact of any immediate precrisis boom or slump, and there is no well-established method for doing this.3 In this work, a linear trend is estimated through the actual output series during a seven-year precrisis period that ends three years before the onset of the crisis. In a number of cases, however, this procedure yielded negative trend growth rates, implying that output per capita would decline indefinitely even in the absence of a crisis. In these cases, the precrisis window was extended to 20 years before the crisis and this period was used instead.

One appeal of this approach is that it is simple, transparent, and easy to implement for a large set of countries. It is important that its linearity also facilitates the decomposition of output losses into the factors of production, namely losses in capital, labor, and total factor productivity. The use of a seven-year horizon allows the analysis to abstract from the immediate postcrisis fluctuations in output and focus on medium-term effects. An even longer horizon, such as a 10-year horizon, would have been preferable, but such a horizon would have limited the ability to study a number of crises that occurred in the late 1990s and early 2000s.

The key stylized facts that emerge from the analysis are sobering: output typically does not recover to its precrisis trend. On average, output falls steadily below its precrisis trend until the third year after the crisis, and does not rebound thereafter (Figure 9.3, panel a). The medium-term output losses following banking crises are substantial: seven years after the crisis, output has declined relative to trend by close to 10 percent on average. As the shaded area measuring the 90 percent confidence band indicates, the average decline relative to trend is statistically significant. To put the losses associated with banking crises in perspective, Figure 9.3 also reports the evolution of output relative to trend following currency crises in panel b. Estimated losses following currency crises are much smaller, about one-third of the average loss associated with banking crises.

Figure 9.3Output Evolution after Banking and Currency Crises

Sources: World Bank, World Development Indicators; and IMF staff calculations.

Note: Figure reports mean difference form year t = –1; 90 percent confidence interval for estimated mean (shaded area); first year of crisis at t = 0.

At the same time, however, the slope of the trend itself does not appear to be affected by the crisis. Although annual growth tends to fall substantially below the precrisis trend during the first two years of the crisis, it is statistically indistinguishable from the precrisis trend thereafter (Figure 9.4, panel a). The above-normal growth required to return output to the previous trend does not tend to materialize. The four-year average of growth ending in the seventh year after the crisis has a mean difference with respect to the precrisis trend growth rate of only −0.2 percentage point per year, with a standard error of 0.5 percentage point.

Figure 9.4Growth Evolution after Banking and Currency Crises

Sources: World Bank, World Development Indicators; and IMF staff calculations.

Note: First year of crisis at t = 0; 90 percent confidence interval for estimated mean (shaded area).

In addition, the variation in outcomes is substantial. For example, although the change in output relative to trend following banking crises has a mean of −10 percent, the middle 50 percent of cases had a range of −26 percent to +6 percent. On average there is no rebound to the precrisis trend, but more than a quarter of cases ultimately exceeded the precrisis trend. Similarly, whereas growth tends to return to the precrisis trend rate, the middle 50 percent of cases had a deviation relative to the precrisis growth trend ranging from of −2.8 percentage points to +1.7 percentage points.

To set these findings against a significant historical benchmark, the same methodology was applied to a measure of global output in the aftermath of the Great Depression.4 Following a stock market crash in the United States in 1929, numerous economies experienced banking crises. Consistent with the cross-country results discussed previously, the level of global output did not return to its precrisis trend in the medium term, and was 27 percent below the precrisis trend by 1936, with the bulk of the decline occurring during 1929–32. At the same time, the global growth rate eventually returned to the precrisis trend (by 1934), in line with the more recent crisis episodes.

The robustness of the results was checked by considering alternative approaches to estimating the precrisis trend. First, the calculations were repeated with the precrisis window ending one year rather than three years before the crisis. Second, the precrisis trend was computed based on a longer precrisis window (from t − 20 to t − 3). Third, the precrisis trend was computed using the real-time growth projections of IMF staff prepared for the spring World Economic Outlook in the year before the crisis. Overall, as reported in Appendix 9A, the output losses obtained using the different approaches were statistically indistinguishable from the baseline. Similarly, the result that the growth rate eventually returns to the precrisis trend is robust to using these alternative precrisis trend measures. To understand better which components of output are adversely affected during banking crises, the chapter now turns to analyzing the underlying factors behind the postcrisis medium-term output dynamics.

Decompositions: Why Does Output Not Recover?

This section decomposes medium-term output losses into factor inputs and demand components to help understand which factors drive them. Learning about the underlying forces could provide insights into the likely evolution of output after the banking crises associated with the 2007–09 global crisis, and what type of policies may be relevant to reduce the ultimate losses. Before presenting the results, the main channels through which banking crises may affect output in the medium term are reviewed.

What Are Possible Effects on the Key Sources of Output?

A useful way to examine why output per capita often does not recover to its precrisis trend is to analyze what happens to the key elements of an economy’s production process—labor inputs (which can be thought of as depending on the employment rate and labor force participation), capital inputs, and total factor productivity. Of course, changes in output components following banking crises could reflect a decline in the productive potential of the economy, but they could also reflect a persistent fall in aggregate demand, although the latter is likely to explain only a small part of the medium-term losses. From a theoretical perspective, banking crises may affect these production components in several ways.

First, the medium-term effect of a crisis on labor force participation is uncertain because there are two opposing forces. On the one hand, grim employment prospects may discourage job seekers and prompt them to leave the labor force, especially if there are incentives to retire early. On the other hand, in times of economic hardship, second-income earners may enter the labor force to help compensate for the loss of family income or wealth.5

Second, the medium-term employment rate would be affected adversely if banking crises lead to an increase in the underlying (so-called structural) unemployment rate. For example, the crisis may imply a need for a substantial reallocation of labor across sectors, which can take time and increase medium-term frictional unemployment. Perhaps more important, the large initial increase in the actual unemployment rate induced by the crisis could persist if rigid labor market institutions (strict employment protection laws, generous unemployment benefits, and the like) complicate the task of finding a new job. Long spells without employment may also impair professional and on-the-job skills, making it even more difficult for the long-term unemployed to find jobs, resulting in so-called hysteresis effects (Blanchard and Wolfers, 2000; Nickell, Nunziata, and Ochel, 2005; and Bassanini and Duval, 2006; among others).

Third, a banking crisis may slow capital accumulation by depressing investment over a protracted period. As the supply of credit becomes more limited, firms face tougher financing conditions in the form of tighter lending standards and higher effective costs of borrowing, and profit rates are likely to suffer (Bernanke and Blinder, 1988; and Bernanke and Gertler, 1989). The ability of firms to borrow and invest may be hampered further if the crisis leads to lower asset prices that weaken corporate balance sheets and erode collateral values (Kiyotaki and Moore, 1997). Investment may also suffer if the crisis leads to a sustained increase in uncertainty and risk premiums.

Finally, the effect on total factor productivity is ambiguous based on theoretical considerations, but likely to be negative. On the negative side, as it recovers from the crisis, the financial system may not be able to allocate loanable funds as productively as before the crisis, particularly if high-risk but high-return projects are discouraged by more cautious lending attitudes. In addition, productivity may also suffer as a result of less innovation, given that research and development spending tends to be cut back in bad times (Guellec and van Pottelsberghe, 2001). Also, high-productivity firms may fail because of lack of financing. On the positive side, however, banking crises may have a cleansing effect on the economy by removing inefficient firms and activities and creating incentives to restructure and improve efficiency (Caballero and Hammour, 1994; and Aghion and Saint-Paul, 1998).6

What Do the Data Show?

Medium-term output losses following banking crises are decomposed into underlying components using the following approach: The starting point is the observation that the logarithm of output per capita is equal to the weighted sum of the logarithms of labor force participation, the employment rate, the capital-to-labor ratio, and total factor productivity. Note that, because of data limitations, the decompositions into factor components are based on a smaller sample of 27 observations.

Applying the same procedure for estimating precrisis trends and attributing output losses to their underlying components allows output losses to be decomposed into losses attributable to changes in the employment rate, labor force participation, the capital-to-labor ratio, or total factor productivity. Specifically, for each output component, the precrisis trend is estimated for the same precrisis period as the output trend. This approach ensures that, based on the assumed Cobb-Douglas production function, the factor input contributions add up exactly to the total output loss.

The decompositions are based on a Cobb-Douglas production function of the form Y = AEαK1–α, where A denotes total factor productivity, E denotes employment, and K denotes the capital stock. The employment share α is assumed to be 0.65. Given the assumption of constant returns to scale, the production function can be expressed in per capita terms by dividing by population, P, yielding

Finally, taking logs, and noting that

in which LF denotes the labor force, yields the decomposition used in the analysis:

in which KE represents the capital-to-labor ratio, ELF is the employment rate, and LFP is the labor force participation rate. Note that because total factor productivity, A, is obtained as the residual from the decomposition, it may reflect errors in the measurement of the factor inputs.

To complement the analysis, an analogous decomposition is performed for the demand-side components of output: investment, consumption, exports, and imports. Note, however, that because the demand components are additive, the losses of the aggregate demand components do not sum exactly to the total output loss. The results for the two types of output loss decompositions are presented in Figure 9.5 and Figure 9.6. For each component of output, the 90 percent confidence bands are reported to indicate the statistical significance of the estimates.

Figure 9.5Output Decomposition

Sources: World Bank, World Development Indicators; and IMF staff calculations.

Note: Figure reports mean difference form year t = –1; 90 percent confidence interval for estimated mean (shaded area); first year of crisis at t = 0.

Figure 9.6Demand-Side Decomposition

Sources: World Bank, World Development Indicators; and IMF staff calculations.

Note: Figure reports mean difference form year t = –1; 90 percent confidence interval for estimated mean (shaded area); first year of crisis at t = 0.

The decomposition shows that the measured medium-term losses in GDP per capita can be attributed to roughly equal losses in three of the four components of output—the employment rate, the capital-to-labor ratio, and total factor productivity (Figure 9.5).7 Regarding total factor productivity, after a significant initial decline, the level gradually moves closer to the precrisis trend toward the end of the seven-year horizon. This is consistent with the notion of labor hoarding that decreases over time. Nevertheless, the medium-term loss in total factor productivity still accounts for about one-third of the total output loss. Its magnitude, however, is not statistically significant seven years after the crisis, although it is significant in the short term. Regarding the other two key components, the initial loss in the employment rate persists into the medium term, while capital losses worsen steadily. The finding of an adverse impact on the capital-to-labor ratio is consistent with the demand-side decompositions that show a large and significant decline in investment of about 30 percent relative to its precrisis trend (Figure 9.6). The consumption loss is also notable and significant, at about 15 percent.

Overall, the decompositions suggest that higher unemployment rates, slower capital accumulation, and lower productivity growth play an important role in explaining medium-term output losses following banking crises. In other words, output per capita does not recover to its precrisis trend because capital per worker, the unemployment rate, and productivity do not typically return to their precrisis trends within seven years of the crisis. This finding suggests that pre- and postcrisis macroeconomic conditions and policies could play a role in shaping medium-term output dynamics—an issue examined in the next section.

What Factors are Associated with Medium-Term Output Losses?

As illustrated in the previous sections, there is substantial variation in medium-term output losses across banking crises. To explain these variations, this section explores how output losses are related to various factors—such as macroeconomic, structural, and policy conditions—both before and after the crisis. The subsequent section explores the relationship between these factors and postcrisis growth losses.

The analysis of pre- and postcrisis factors proceeds in two steps. First, the results of small-scale ordinary least squares (OLS) regressions that consider several factors at a time are presented. These small-scale regressions typically include one or two variables of interest in addition to key control variables. The robustness of the results is then explored using a large-scale OLS regression that considers all of the factors simultaneously.

In addition to the OLS regressions, Bayesian model averaging (BMA) is also used, which allows an examination of whether the associations found for each variable are robust to including additional controls in all the possible ways that those additional controls can be added. The procedure summarizes the results obtained across all possible specifications using two key statistics: the average coefficient value obtained for each variable, and the probability that each variable is statistically “effective” and should be used to predict output losses. A conventional approach in the BMA literature is to refer to a variable as “effective” if its estimated inclusion probability is greater than 50 percent.8 BMA is particularly useful in this investigation because theory is not sufficiently explicit about which variables should be included in the “true” regression. At the same time, however, BMA has substantial data requirements that reduce by half the number of available observations for this analysis. Thus, both the small-scale results (based on a broad sample) and the larger-scale models (based on a restricted sample) are used.

Do Initial Conditions Help Predict Medium-Term Output Losses?

What are the precrisis factors that might explain the magnitude of the eventual output losses? The analysis examines the importance of a range of macroeconomic, structural, and policy environment variables. The sources of the data are reported in Appendix 9A.

The precrisis output position (which identifies the starting position of output relative to trend) and the initial change in output during the first year of the crisis (which indicates the severity of the crisis in the short term) are potentially important control variables. Both the small-scale OLS results and the BMA analysis indicate that the severity of the crisis, measured by the first-year change in output, has strong predictive power for medium-term output losses (Table 9.1, row 20). A 1 percentage point fall in output relative to trend in the first year of the crisis is associated with a 1.1 to 1.8 percentage point gap between output and the precrisis trend by year t + 7. This result underscores the notion that banking crises have long-lasting effects on output. At the same time, a depressed level of output relative to trend before the crisis appears to carry over, and is associated with a significantly larger medium-term output loss (Table 9.1, row 19).9 Based on these results, the two initial output variables are included as controls in all the remaining regressions.

Table 9.1.Output Losses versus Initial Conditions(Dependent variable: output at t + 7 in percent of precrisis trend)
(1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)
(1)Investment/GDP−0.989***−1.211***−1.602
[-3.120][-2.825](1.000)
(2)Investment/GDP gap0.335−1.049−0.388
[0.889][-1.671](0.381)
(3)Current account/GDP0.765**0.06320.000
[2.016][0.167](0.000)
(4)Current account/GDP gap0.9640.5250.189
[1.593][0.571](0.196)
(5)Inflation0.1160.00535−0.002
[1.500][0.0632](0.042)
(6)Inflation gap−0.196**−0.0627−0.032
[-2.243][-0.475](0.258)
(7)Fiscal balance0.501−0.5410.000
[1.205][-1.102](0.000)
(8)Fiscal balance gap1.256**0.4800.013
[2.042][0.796](0.022)
(9)Real exchange rate gap−0.176
[-1.274]
(10)Real interest rate gap−0.127
[-0.166]
(11)Log (PPP GDP per capita)0.01750.02830.000
[0.736][0.635](0.000)
(12)Credit/GDP−0.152−0.03160.005
[-1.616][-0.299](0.073)
(13)Credit/GDP gap0.2040.4380.027
[0.503][0.993](0.109)
(14)Currency crisis−0.141*−0.155−0.082
[-1.878][-1.483](0.558)
(15)U.S. Treasury bill rate0.5431.0110.026
[0.528][0.999](0.038)
(16)External demand shock−0.100−0.113*−0.012
[-1.200][-1.960](0.089)
(17)Financial openness/GDP0.059***0.007500.002
[3.031][0.499](0.094)
(18)Trade openness/GDP−0.133−0.02970.000
[-1.549][-0.421](0.000)
(19)Precrisis output1.601***1.328***1.598***1.027***0.950***1.425**1.538***0.900***1.685***1.632***0.751**0.9010.916
[3.844][3.875][4.855][2.691][3.174][2.435][3.639][2.700][3.931][3.807][2.175][1.437](0.871)
(20)First-year output change1.681***1.583***1.573***1.781***1.841***1.0691.752***1.665***1.552***1.699***1.799***1.289***1.175
[3.051][3.551][3.608][3.406][3.547][0.992][3.039][3.280][2.694][3.046][3.271][3.379](1.000)
(21)Constant term0.0558**0.162**−0.01810.0929**−0.0511*−0.0662−0.0771*−0.02140.0451*−0.0863−0.04860.1250.337
[-2.652][2.156][-0.726][-2.759][-1.970][-1.182][-2.036][-0.806][-2.003][-1.271][-1.159][0.791](1.000)
Number of observations88858087812688778888524444
R-squared0.3340.4080.4090.3340.3690.2560.3380.2950.3530.3390.3140.763
Source: IMF staff calculations.Note: PPP = purchasing power parity. Columns 1–12 report estimation results based on ordinary least squares with robust t-statistics in square brackets. The symbols ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively. Column 13 reports estimation results based on Bayesian model averaging with tha estimated probability of inclusion of each variable in parentheses. The term “gap” denotes the deviation of the variable from its precrisis historical average (years t-10 to t-17, where t denotes the crisis year) during the last three years preceding the crisis.
Source: IMF staff calculations.Note: PPP = purchasing power parity. Columns 1–12 report estimation results based on ordinary least squares with robust t-statistics in square brackets. The symbols ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively. Column 13 reports estimation results based on Bayesian model averaging with tha estimated probability of inclusion of each variable in parentheses. The term “gap” denotes the deviation of the variable from its precrisis historical average (years t-10 to t-17, where t denotes the crisis year) during the last three years preceding the crisis.

The prominent role of investment and capital losses would suggest that the level and evolution of precrisis investment would be good predictors of the eventual output losses. Regression results provide strong evidence that countries with high precrisis investment-to-GDP ratios, measured as the average investment-to-GDP ratio during the three years before the crisis, tend to have large output losses (Table 9.1, row 1; Figure 9.7). In contrast, the investment gap, defined as the deviation of the investment-to-GDP ratio during the last three years from its historical average, is not statistically significant (Table 9.1, row 2).10 Potential interpretations of these results are returned to later in the section. Nevertheless, the precrisis investment share result is particularly robust and holds even after controlling for the level of the current account balance. This outcome suggests that countries with high investment rates tend to experience larger output declines following banking crises, irrespective of whether the investment is financed by foreign or domestic savings.

Figure 9.7Output Evolution and Precrisis Investment Share

Sources: World Bank, World Development Indicators; and IMF staff calculations.

Note: Figure reports mean difference from year t = –1 for countries with precrisis investment share below median (panel a) and above median (panel b); 90 percent confidence interval for estimated mean (shaded area); first year of crisis at t = 0.

By limiting the room for policy maneuvering, the buildup of macroeconomic imbalances may also imply higher medium-term output losses after a crisis. The analysis considers the precrisis levels and dynamics of several variables—such as inflation, the current account balance, the fiscal balance, the real exchange rate, the real interest rate—that may capture the notion of macroeconomic imbalances.11 The evidence is mixed that rising imbalances, and by implication, more limited policy space that would constrain the ability of countries to run countercyclical macroeconomic policies, are associated with larger output losses. In particular, the results based on the small-scale regressions suggest that countries with larger current account deficits, rising inflation, and a deteriorating fiscal balance before the crisis experienced significantly larger output losses (Table 9.1, rows 3, 6, 8). But the BMA analysis (Table 9.1, column 13) suggests that the evidence is strong only for rising inflation before the crisis. It is important to bear in mind that more policy space does not necessarily mean that it was used—an issue returned to later.12

An interesting finding is that postcrisis output losses are not significantly correlated with the level of income (Table 9.1, row 11). In fact, the evolution of output after banking crises for high-income, middle-income, and low-income countries is similar (Figure 9.8). This finding is consistent with the notion that banking crises represent an “equal opportunity menace” (Reinhart and Rogoff, 2009b) for countries across the income distribution. At the same time, there is mixed evidence that a higher precrisis level of financial development is associated with larger output losses (Table 9.1, row 12).13

Figure 9.8Output Evolution and Precrisis Income Level

Sources: World Bank, World Development Indicators; and IMF staff calculations.

Note: Figure reports mean difference from year t = –1 by quartile of real purchasing-power-parity GDP per capita; 90 percent confidence interval for estimated mean (shaded area); first year of crisis at t = 0.

Currency crises that coincide with banking crises, so-called twin crises, are robustly associated with larger output losses (Table 9.1, row 14). The results for the openness indicators, however, are mixed (Table 9.1, row 17 and row 18). The small-scale regression approach suggests that financial openness is associated with smaller losses, and is consistent with work that finds that deeper financial integration reduces the risk of a sudden stop in capital flows, and enhances the ability to smooth spending (Calvo, Izquierdo, and Mejia, 2008; and Abiad, Leigh, and Mody, 2009). However, the evidence is weaker based on the broader specification. Evidence for trade is even weaker. Turning to external conditions, the level of the U.S. Treasury bill rate before the crisis is not found to be a significant predictor of output losses (Table 9.1, row 15). The evidence that an adverse external demand shock occurring at the time of the banking crisis is correlated with larger output losses is mixed (Table 9.1, row 16).14

Finally, the precrisis levels of various structural policy reform indicators are not significantly correlated with medium-term output losses, and are not presented in Table 9.1. The discussion returns to the possible role of structural policies in the next section, which considers whether countries that undertook structural reforms following the crisis experienced smaller output losses.15

What key points should be taken away from these regression results? The empirical analysis suggests that the first-year loss is important in predicting the eventual output losses following a banking crisis. This outcome is consistent with the notion that output dynamics are especially persistent following large shocks. What could explain this? A possible explanation is that bankruptcies lead to fire sales of capital assets that have significant sunk costs and take time to rebuild. Also, an impaired financial system may need time to heal and intermediate financial capital effectively, and labor and product market rigidities could impede the necessary reallocation of labor and capital following a crisis. These interpretations are consistent with the finding that all factors of production contribute to medium-term output losses.

Related to the dynamics of capital accumulation, the finding that the precrisis investment rate is a robust predictor of the postcrisis output loss is particularly striking. This finding, together with the earlier result that investment and capital deepening decline in the medium term following banking crises, is consistent with a number of potential interpretations.16 In some cases, it may be that the output loss reflects the unwinding of excessive investment built up over a protracted period. To the extent that some investment during the precrisis period was wasteful, output losses may have taken place even without a crisis, but gradually. However, a full investigation into the underlying reasons for the remarkably strong correlation between the precrisis investment level and medium-term output losses is an issue that merits further investigation but is beyond the scope of this chapter.

After the Crisis: What Is Associated with Smaller Output Losses?

What role do policies have in mitigating the ultimate output loss after the crisis has hit? It is important to acknowledge that the following discussion seeks to identify patterns rather than establish causality between postcrisis output evolution and policies. As discussed in the literature, the two-way relationship between postcrisis policies and outcomes complicates any causal inference. For example, is it that financial reform during or after a banking crisis leads to increased financial intermediation and a lower output loss? Or, that a lower output loss leads to higher demand and thus higher financial intermediation and also gives the authorities the policy space to implement important financial sector reforms? These difficult questions cannot be answered within this chapter’s regression framework.

The discussion focuses on domestic macroeconomic policies and structural reforms, and on external conditions and policies abroad. As in the analysis of precrisis factors, this section present the regression results (Table 9.2) based on both full-sample OLS and restricted-sample BMA analysis. As before, all regressions control for key initial output variables.

Table 9.2.Output Losses versus Postcrisis Conditions and Policies(Dependent variable: output at t + 7 in percent of precrisis trend)
(1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)
(1)Real government consumption growth0.202**0.244*0.405**0.263
[2.520][1.843][2.264](0.648)
(2)Change in real interest rate−0.0850−0.493**−0.580−0.530
[-0.404][-2.280][-1.577](0.708)
(3)Real exchange rate appreciation0.135*−0.0105−0.418*−0.038
[1.785][-0.0753][-2.047](0.166)
(4)Change in capital account liberalization0.166***0.147**0.02970.007
[4.627][2.290][0.433](0.085)
(5)Change in financial liberalization index0.108**0.01700.149*0.002
[2.583][0.302][1.769](0.044)
(6)Change in trade liberalization index−0.0456−0.0632−0.122−0.013
[-0.950][-1.123][-1.506](0.149)
(7)Change in government efficiency index−0.005000.01320.129*0.078
−0.0774[0.213][2.044](0.608)
(8)U.S. Treasury bill rate−1.4040.490−4.459−2.820
[-1.012][0.178][-1.524](0.400)
(9)External demand shock−0.960***−1.161−1.073−0.415
[-3.156][-1.611][-1.668](0.411)
(10)Precrisis output1.213***1.038***1.371***1.079***0.997***1.384***1.162***1.601***1.753***1.137***1.124***0.9070.143
[4.666][2.791][4.292][3.537][4.358][4.456][2.398][3.783][4.427][3.453][3.061][1.687](0.184)
(11)First-year output change2.032***2.107***1.750***2.191***2.262***2.145***1.749***1.714***1.875***2.365***2.220***3.136***2.693
[3.396][2.941][2.884][3.560][3.529][3.526][2.591][3.158][3.558][2.667][3.330][2.889](1.000)
(12)Constant term0.0560*0.0470*−0.03410.0929**0.0882**−0.0198−0.05350.0227−0.00409−0.0366−0.0789*0.06390.052
[-2.065][-2.059][-1.471][-4.010][-3.510][-0.869][-1.485][0.284][-0.177][-0.260][-1.964][0.385](1.000)
Number of observations77597465657853888850493030
R-squared0.3980.2830.3420.4590.3970.3880.2810.3440.3960.5060.4500.709
Source: IMF staff calculations.Note: Columns 1-12 report estimation results based on ordinary least squares with robust t-statistics in square brackets. The symbols ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively. Column 13 reports estimation results based on Bayesian model averaging with the estimated probability of inclusion of each variable in parentheses. Structural reform variables (trade, financial capital account, and government efficiency) measure change in index from t to t + 7.
Source: IMF staff calculations.Note: Columns 1-12 report estimation results based on ordinary least squares with robust t-statistics in square brackets. The symbols ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively. Column 13 reports estimation results based on Bayesian model averaging with the estimated probability of inclusion of each variable in parentheses. Structural reform variables (trade, financial capital account, and government efficiency) measure change in index from t to t + 7.

Short-term demand management policies (fiscal and monetary) implemented after the crisis has hit may play a role both in reducing the size of the initial output loss and in aiding the recovery. To measure changes in discretionary fiscal policy, the analysis follows the approach of IMF (2009a) and uses the growth in real government consumption. Given data availability, the monetary policy stance is measured as the change in real lending rates. In both cases, to capture the short-term response of macroeconomic policies, the variables are computed for the crisis year and the subsequent three years. The variables are designed to measure a notion of stimulus (rather than policy space), and thus differ from those used in the precrisis analysis. The findings are that a stronger short-term fiscal policy response (a larger increase in government consumption) is significantly associated with smaller medium-term output losses (Table 9.2, row 1).17 The evidence on the monetary policy stance is mixed, possibly reflecting a weaker monetary policy transmission mechanism after banking crises. A decline in real interest rates is associated with smaller output losses, but only in some specifications (Table 9.2, row 2). There is also mixed evidence that real exchange rate depreciations are associated with smaller output losses (Table 9.2, row 3).

Advancing structural reforms may also play a role in boosting output during the postcrisis period. The exercise considers reform efforts in several areas, such as domestic financial reform, capital account and trade liberalization, and structural fiscal reforms. In each case, the reform effort is measured as the change in various indices mentioned earlier during the postcrisis period (rather than the levels that were used in the precrisis analysis).18 Overall, the evidence that structural reform efforts are significantly associated with smaller output losses is mixed. Liberalization of the capital account is highly correlated with smaller output losses in small-scale regressions, although its statistical significance declines when considered in larger-scale frameworks (Table 9.2, row 4). Domestic financial reforms are also significantly positively associated with output losses in small-scale regressions, but less so in larger-scale frameworks (Table 9.2, row 5). Trade liberalization is not significantly related to output losses (Table 9.2, row 6). Finally, there is some positive evidence on the link between improvements in government efficiency and output losses, although the increased significance of this structural variable in the broader specifications appears to be partly due to the change in the sample composition (the number of observations drops to 30).

Finally, policies and conditions abroad may also be important in reducing output losses by improving the external environment during the postcrisis period. The results indicate that larger domestic output losses are significantly related to the occurrence of adverse external demand shocks during the postcrisis period (Table 9.2, row 9). In addition, there is some evidence that larger output losses are significantly associated with higher global short-term interest rates (Table 9.2, row 8).19

How should these empirical findings be interpreted? Overall, the findings suggest that expansionary short-term macroeconomic policies are associated with smaller medium-term output losses. This is consistent with the notion that countercyclical fiscal and monetary policies may help dampen path-dependence effects by cushioning the downturn after the crisis, which carry over into smaller measured output losses in the medium term.

The relationship between postcrisis structural policy reforms and output losses is somewhat weaker. However, this could be due to well-known difficulties in measuring the timing, magnitude, and sequencing of structural reforms,20 as well as the possibility that structural reforms and capacity building may take a longer time to bear fruit by increasing output. At the same time, the spillover effects of global conditions may be important, given the strong association between the external environment and the eventual output losses.

Overall, the regression analysis provides suggestive evidence that domestic fiscal and monetary stimulus and favorable global conditions may mitigate medium-term output losses. There is also some mixed evidence on the beneficial role of structural policy reform. However, much can still be learned about the processes and interactions associated with output losses following banking crises.

Is this Time Different?

So far, this chapter has assessed macroeconomic performance after previous banking crises—those before the 2007–09 global financial crisis. A natural question is whether, in economies that suffered banking crises during 2007–09, output has evolved in line with the historical pattern found for earlier crises.

The analysis now estimates medium-term output losses for the 25 economies that entered banking crises during 2007–09 and compares them with those of previous banking crises.21 Exactly the same approach as described above is used here.

The results suggest that the most recent banking crisis economies have experienced output losses that are larger than the historical average. As of 2012, output is, on average, about 17 percent below the precrisis trend (Figure 9.9), which compares with an average output loss for previous crises after five years of 9 percent. In addition, some evidence suggests that the output losses may be longer lasting. About one-third of previous banking crises saw output return to or exceed the precrisis trend within five years. In contrast, as of 2012, none of the 25 recent banking crisis economies (except Mongolia) have seen output return to or exceed the precrisis trend. Overall, therefore, the output losses appear to be larger than in the past, and may be more persistent.

Figure 9.9Output Evolution after the 2007–09 and Historical Banking Crises

Sources: World Bank, World Development Indicators; and IMF staff calculations.

Note: Figure reports mean difference from year t = –1; 90 percent confidence interval for estimated mean (shaded area); first year of crisis at t = 0.

Conclusions

Based on an investigation of 88 historical banking crises in a wide range of economies, this chapter finds that economic activity tends to contract sharply following such crises, with no rebound to the precrisis trend in the medium term. On average, seven years after the crisis, output has declined by about 10 percent relative to the precrisis trend. The above-average growth needed to return output to the previous trend does not tend to materialize, although growth does eventually return to its precrisis rate. Also, the depressed path of output tends to result from reductions of roughly equal proportions in the employment rate, the capital-to-labor ratio, and total factor productivity.

A large variation in outcomes across countries is also found, with output eventually returning to or even exceeding the precrisis trend in one-third of cases. An exploration of this variation in outcomes finds that medium-term output losses are strongly correlated with conditions at the onset of the crisis, including the severity of the initial contraction in activity, the occurrence of twin banking and currency crises, and a high precrisis level of investment. Short-term fiscal and monetary stimulus is also associated with smaller medium-term output losses. There is also some evidence that structural policy reforms implemented after the onset of the crisis can play a supportive role. Although the contemporaneous nature of these variables complicates a causal interpretation, the results are consistent with the notion that policies implemented in the aftermath of a crisis can help to limit the medium-term damage.

Finally, the performance of economies that experienced banking crises during 2007–09 bears a strong resemblance to previous banking crises. These economies have suffered even deeper output losses, as of 2012 averaging 17 percent relative to the precrisis trend. Moreover, there is little evidence of output returning to the precrisis trend. The greater output losses may reflect the highly synchronized nature of the recent episodes, as part of the global financial crisis, and in some cases, the additional role of sovereign debt problems. We leave a further investigation of the recent crises to future research.

Appendix 9A. Data and Robustness Results

This appendix provides details on the data used in the analysis, a list of the 88 banking crisis episodes considered in the analysis (Table 9A.1), and the results of robustness exercises on measuring output losses.

Table 9A.1List of Historical Banking Crisis Episodes
CountryYearOutput loss
Algeria1990−14
Argentina1980−15
Argentina198919
Argentina1995−12
Argentina20013
Bangladesh198713
Benin1988−11
Bolivia19861
Bolivia19945
Brazil1990−12
Brazil19943
Burkina Faso19901
Burundi1994−36
Cameroon1987−86
Cameroon19959
Central African Rep.1976−18
Central African Rep.199518
Chad198340
Chad1992−35
Chile19766
Chile19817
Colombia1982−13
Colombia1998−15
Congo, Dem. Rep. of1983−8
Congo, Dem. Rep. of1991−69
Congo, Dem. Rep. of1994−32
Congo, Republic of1992−37
Costa Rica198711
Costa Rica19943
Côte d’lvoire1988−19
Ecuador1982−32
Ecuador19988
Egypt1980−12
El Salvador198938
Finland1991−15
Ghana1982−2
Guinea19936
Guinea-Bissau1995−39
Haiti1994−11
India19933
Indonesia1997−47
Israel1977−34
Jamaica1996−32
Japan1997−12
Jordan1989−29
Kenya1985−8
Kenya1992−16
Korea1997−23
Kuwait198223
Liberia1991−10
Madagascar19885
Malaysia1997−38
Mali1987−8
Mauritania1984−10
Mexico1981−28
Mexico19949
Morocco1980−26
Nepal198813
Nicaragua199018
Nicaragua20007
Niger1983−44
Nigeria19913
Norway19910
Panama19880
Paraguay1995−26
Peru1983−42
Philippines1983−33
Philippines199717
Senegal1988−7
Sierra Leone1990−54
Spain1977−37
Sri Lanka19892
Swaziland1995−23
Sweden1991−14
Thailand19838
Thailand1997−52
Togo19938
Tunisia199116
Turkey19821
Turkey20007
Uganda199418
United States19880
Uruguay1981−31
Uruguay200211
Venezuela1994−7
Vietnam19973
Zambia199517
Zimbabwe1995−21
Note: Table reports first year of banking crisis, and output loss in percent of precrisis trend seven years after the crisis (log points).
Note: Table reports first year of banking crisis, and output loss in percent of precrisis trend seven years after the crisis (log points).

Data Sources

The main data sources for this chapter are the IMF’s World Economic Outlook (WEO) and International Financial Statistics (IFS) databases, and the World Bank’s World Development Indicators (WDI). Additional data sources are listed below.

Data on real GDP and its demand components are from the WDI, and are spliced with WEO data for observations after 2007 for which WDI data are unavailable. The current account balance, the GDP deflator, and the fiscal balance are also taken from the WEO, while the exchange rate series are taken from the IFS. The domestic real interest rate is defined as the difference between the nominal lending rate, taken from the IFS, and GDP-deflator inflation.

For the growth accounting exercises, the capital stock data are taken from Bosworth and Collins (2003). For observations not included in the Bosworth and Collins data set, the capital stock is constructed using the perpetual inventory method, with a depreciation rate of 5 percent, and using real investment data. The employment and labor force data come from the WEO.

Financial development is measured using the ratio of bank credit to GDP, following Abiad, Dell’Ariccia, and Li (2011). Bank credit to the private nonfinancial sector is taken from the IFS. Breaks in these data are identified using the IFS Country Notes publication, and data are growth-spliced at these points.

Financial openness is calculated as the sum of foreign assets and foreign liabilities divided by GDP, using the External Wealth of Nations Mark II database of Lane and Milesi-Ferretti (2006). Trade openness is defined as the sum of exports and imports divided by GDP. Partner-country growth, used to compute the external demand shocks, is taken from the WEO, and the three-month U.S. Treasury bill rate is obtained from Thomson Reuters Datastream.

The structural reform indicators measuring trade liberalization, capital account liberalization, financial liberalization, and government efficiency come from the IMF, and are described in greater detail in Giuliano, Mishra, and Spilimbergo (2009) and IMF (2008).

Robustness: Alternative Measures of Output Losses

The baseline measure of the output loss is compared with three alternative measures based on the following different versions of the precrisis trend.

  • Alternative 1. Precrisis window ending one year before crisis. The precrisis trend is computed as in the baseline, except that the estimation window for the precrisis trend ends one year before the crisis, rather than three years before the crisis as it does in the baseline.

  • Alternative 2. Longer estimation window. The estimate of the precrisis trend is obtained based on a longer precrisis window going back 20 years before the crisis and ending 3 years before the crisis.

  • Alternative 3. Precrisis trend based on real-time IMF staff forecasts. The output losses were recomputed using the real-time medium-term growth projections of IMF staff prepared for the spring World Economic Outlook in the year before the crisis. In particular, the precrisis trend growth rate is defined as the IMF country desk forecast for real GDP growth in year t = 4 made in year t = −1, where t = 0 is the year of the crisis. Note that these real-time forecasts were only available for the post-1989 period.

Overall, as Figure 9A.1 reports, the losses obtained using the different approaches were highly correlated, and all confirm the finding of large and statistically significant output losses after banking crises. The 90 percent confidence bands for each measure overlap with the baseline measure.

Figure 9A.1Robustness: Alternative Precrisis Trends in Output Evolution after Banking Crises

Sources: World Bank, World Development Indicators; and IMF staff calculations.

Note: Figure reports the estimated mean difference from year t = –1; 90 percent confidence interval for estimated mean (shaded area); first year of crisis at t = 0.

References

    AbiadA. G. Dell’Ariccia and B. Li2011Creditless RecoveriesIMF Working Paper 11/58 (Washington: International Monetary Fund).

    AbiadAbdulDanielLeigh and AshokaMody2009Financial Integration, Capital Mobility, and Income ConvergenceEconomic Policy Vol. 24 No. 58 pp. 241305.

    AghionPhilippe and GillesSaint-Paul1998On the Virtue of Bad Times: An Analysis of the Interaction between Economic Fluctuations and Productivity GrowthMacroeconomic Dynamics Vol. 2 No. 3 pp. 32244.

    AngkinandApanard P.2008Output Loss and Recovery from Banking and Currency Crises: Estimation Issues” (Springfield, Illinois: University of Illinois).

    BassaniniAndrea and RomainDuval2006The Determinants of Unemployment across OECD Countries: Reassessing the Role of Policies and InstitutionsOECD Economic Studies Vol. 2006 No. 1 pp. 786.

    BassaniniAndrea and RomainDuval and Alan S. Blinder1988Credit, Money, and Aggregate DemandAmerican Economic Review Vol. 78 No. 2 (May) pp. 43539.

    BernankeBen S. and MarkGertler1989Agency Costs, Net Worth, and Business FluctuationsAmerican Economic Review Vol. 79 No. 1 (March) pp. 1431.

    BernankeBen S. and MarkGertler1995Inside the Black Box: The Credit Channel of Monetary Policy TransmissionJournal of Economic Perspectives Vol. 9 No. 4 (Autumn) pp. 2748.

    BlanchardOlivier and JustinWolfers2000The Role of Shocks and Institutions in the Rise of European Unemployment: The Aggregate EvidenceEconomic Journal Vol. 110 No. 462 (March) pp. 133.

    BordoMichael2006Sudden Stops, Financial Crises, and Original Sin in Emerging Countries: Déjà vu?NBER Working Paper No. 12393 (Cambridge, Massachusetts: National Bureau of Economic Research).

    BosworthBarry P. and Susan M. Collins2003The Empirics of Growth: An UpdateBrookings Papers on Economic Activity Vol. 2 (2003) pp. 113206.

    BoydJohn H. SungkyuKwak and BruceSmith2005The Real Output Losses Associated with Modern Banking CrisesJournal of Money Credit and Banking Vol. 37 No. 6 pp. 97799.

    CaballeroRicardo and MohammedHammour1994The Cleansing Effect of RecessionsAmerican Economic Review Vol. 84 No. 5 pp. 135068.

    CalvoGuillermo A. AlejandroIzquierdo and Luis-FernandoMejía2008Systemic Sudden Stops: The Relevance of Balance-Sheet Effects and Financial IntegrationNBER Working Paper No. 14026 (Cambridge, Massachusetts: National Bureau of Economic Research).

    CerraValerie and SwetaSaxena2008Growth Dynamics: The Myth of Economic RecoveryAmerican Economic Review Vol. 98 No. 1 pp. 43957.

    DalyMaryBartHobijn and JoyceKwok2009Labor Supply Response to Changes in Wealth and CreditFRBSF Economic LetterJanuary30.

    FurceriDavide and AnnabelleMourougane2009The Effect of Financial Crises on Potential Output: New Empirical Evidence from OECD countriesOECD Economics Department Working Paper No. 699 (Paris: Organization for Economic Cooperation and Development).

    GiulianoPaolaPrachiMishra and AntonioSpilimbergo2009Democracy and ReformsIZA Discussion Paper No. 4032 (Bonn: Institute for the Study of Labor).

    GuellecDominique and Bruno vanPottelsberghe de la Potterie2008R&D and Productivity Growth: Panel Data Analysis for 16 OECD CountriesOECD Economic Studies Vol. 2001 No. 2 pp. 10325.

    GuptaPoonamDeepakMishra and RatnaSahay2007Behavior of Output during Currency CrisesJournal of International Economics Vol. 72 No. 2 pp. 42850.

    HaughDavidPatriceOllivaud and DavidTurner2009The Macroeconomic Consequences of Banking Crises in OECD CountriesOECD Economics Department Working Paper No. 683 (Paris: Organization for Economic Cooperation and Development).

    HoetingJennifer A. DavidMadiganAdrian E. Raftery and Chris T. Volinsky1999Bayesian Model Averaging: A TutorialStatistical Science Vol. 14 No. 4 pp. 382401.

    HutchisonMichael and IlanNoy2002How Bad Are Twins? Output Costs of Currency and Banking CrisesJournal of Money Credit and Banking Vol. 37 No. 4 pp. 72552.

    International Monetary Fund (IMF)2008Structural Reforms and Economic Performance in Advanced and Developing Countries” (Washington: International Monetary Fund). http://www.imf.org/external/np/res/docs/2008/0608.htm.

    International Monetary Fund (IMF)2009aFrom Recession to Recovery: How Soon and How Strong?” in World Economic OutlookApril (Washington: International Monetary Fund).

    International Monetary Fund (IMF)2009bWhat’s the Damage? Medium-Term Output Dynamics after Financial Crisis” in World Economic OutlookOctober (Washington: International Monetary Fund).

    KiyotakiNobuhiro and JohnMoore1997Credit CyclesJournal of Political Economy Vol. 105 No. 2 (April) pp. 21148.

    LaevenLuc and FabianValencia2008Systemic Banking Crises: A New DatabaseIMF Working Paper 08/224 (Washington: International Monetary Fund).

    LaevenLuc and FabianValencia2012Systemic Banking Crises Database: An UpdateIMF Working Paper 12/163 (Washington: International Monetary Fund).

    LanePhilip R. and Gian MariaMilesi-Ferretti2006External Wealth of Nations Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities, 1970–2004Journal of International Economics Vol. 73 No. 2 pp. 22350.

    MasanjalaWinford H. and ChrisPapageorgiou2008Rough and Lonely Road to Prosperity: A Reexamination of the Sources of Growth in Africa Using Bayesian Model AveragingJournal of Applied Econometrics Vol. 23 No. 5 pp. 67182.

    Milesi-FerrettiGian Maria and AssafRazin1998Current Account Reversals and Currency Crises: Empirical RegularitiesNBER Working Paper No. 6620 (Cambridge, Massachusetts: National Bureau of Economic Research).

    NickellStephenLucaNunziata and WolfgangOchel2005Unemployment in the OECD since the 1960s: What Do We Know?Economic Journal Vol. 115 No. 500 pp. 127.

    Pisani-FerryJean and Bruno vanPottelsberghe2009Handle with Care! Post-Crisis Growth in the EUBruegel Policy Brief No. 2009/02 (Brussels: Bruegel).

    ReinhartCarmen and KennethRogoff2009aThe Aftermath of Financial CrisesAmerican Economic Review Vol. 99 No. 2 pp. 46672.

    ReinhartCarmen and KennethRogoff2009bBanking Crises: An Equal Opportunity MenaceNBER Working Paper No. 14587 (Cambridge, Massachusetts: National Bureau of Economic Research).

    SchumpeterJoseph1942Capitalism Socialism and Democracy (New York: Harper1975; originally published in 1942).

Studies that examine the short-term effects of financial crises include, for example, Hutchison and Noy (2002), Borda (2006), Gupta, Mishra, and Sahay (2007), Haugh, Ollivaud, and Turner (2009), and IMF (2009a).

“Transition countries” are defined based on the classification in the IMF World Economic Outlook of May 1993.

See Angkinand (2008) for a review of alternative methods for estimating output losses associated with a crisis.

For this exercise, an aggregate purchasing-power-parity-weighted real GDP was constructed of a broad sample of countries going back more than 100 years, with the help of the Maddison (2003) Historical Statistics database.

There is some evidence suggesting that the additional-worker effect may have played a role in the 2007–09 crisis, because the female participation rate rose as the male participation rate fell in the United States (Daly, Hobijn, and Kwok, 2009).

The underlying concept of “creative destruction” was first introduced by Schumpeter (1942).

The contribution of labor force participation is positive, albeit small and statistically insignificant.

For additional details on BMA, see, for example, Hoeting and others (1999), and Masanjala and Papageorgiou (2008).

Note that, in the three years before a banking crisis episode, the level of output is, on average, below its trend, suggesting that banking crises are not typically preceded by a precrisis boom. In the sample of 88 banking crises, the average deviation is about −3 percent.

The precrisis historical average level is based on the seven-year period ending three years before the crisis.

The dynamics are captured by considering the deviations of these variables from their country-specific historical averages during the precrisis period (the gaps). Using country-specific averages allows for the possibility that different countries may have different explicit or implicit inflation targets or fiscal rules. For example, a 3 percent inflation rate may provide less space for monetary easing in a country with inflation normally at 1 percent than in a country with an inflation norm of 5 percent. For each variable, the gap value is constructed as a deviation of the average precrisis value (from t − 3 to t − 1) from the country-specific average value (from t − 10 to t − 3). Using government debt to measure fiscal space was not possible for the sample of countries considered here because of limited data availability.

Two other domestic policy variables—the real exchange rate and the real interest rate before the crisis, measured relative to their historic averages—do not appear to have predictive power for medium-term output losses (Table 9.1, row 9 and row 10).

The analysis also considers whether an increase in the credit-to-GDP ratio relative to each country’s own historical average level (the credit-to-GDP gap) plays a role, finding it to be statistically insignificant. The question of whether there is a nonlinear link between the level of financial deepening and output losses is left for further research.

The external demand shock is measured as a dummy variable that equals 1 in year t whenever partner-country growth from year t through t + 4 is in the lowest 5 percent of the entire sample. Partner-country growth is defined as the per capita output growth of a country’s trading partners weighted by their shares in the country’s total exports.

The analysis draws on the database of structural reforms prepared by the Research Department of the IMF. The database covers 150 advanced and developing countries and eight sectors. This chapter uses the domestic financial sector reform index (which includes measures of securities markets and banking sector reforms) and the capital account liberalization index (which summarizes a broad set of restrictions), the trade liberalization index (based on average tariffs), and the fiscal sector reform index (based on tax rates and the efficiency of revenue collection and public spending). It also uses various measures of labor market flexibility, including on employment protection, unemployment benefit replacement ratios, and tax wedges. See IMF (2008) and Giuliano, Mishra, and Spilimbergo (2009) for more details. The indices for product market reforms were not used in the analysis because of insufficient data coverage.

Note that the correlation between the precrisis investment share and the medium-term output loss is largely a reflection of large postcrisis investment losses. In particular, additional regression results not reported here reveal that although a large precrisis investment share is strongly correlated with medium-term investment losses, it is only weakly correlated with medium-term consumption and export losses.

The results imply that raising government consumption by 1 percent of GDP is associated with a reduction in the medium-term output loss of about 1.5 percentage points. The change in government consumption, rather than the change in tax revenue or the fiscal balance, is used as a measure of fiscal stimulus because it lessens reverse-causality concerns. Measuring fiscal stimulus based on the change in tax revenue or the change in the fiscal balance would cause difficulties. A larger deterioration in output implies a greater deterioration in tax revenue and the fiscal balance, complicating the interpretation of the regression coefficients. As expected, repeating the analysis using the change in the fiscal balance yielded a regression coefficient that was statistically indistinguishable from zero.

Regarding labor market liberalization indicators, data availability is limited for the sample of banking crisis countries. Moreover, when data are available, there is often little change post crisis. For these reasons, results for postcrisis labor market indicators are not reported.

Unlike in the small-scale regressions, the global interest rate is significantly related to output losses in the large-scale OLS regression and has a relatively high probability of inclusion (0.63) in the BMA framework.

Note that measurement error in the structural reform indicators will bias the regression coefficients toward zero, making it more difficult to find that the results are statistically significant. Also, the size of the bias depends directly on the magnitude of the measurement error, which is likely to be much larger for unobserved structural reform indicators (such as labor market flexibility or financial sector reform) than for macroeconomic variables (such as government consumption or interest rates).

The 25 systemic banking crises are those that Laeven and Valencia (2012) identify as having begun during 2007–09: Austria, Belgium, Denmark, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Kazakhstan, Latvia, Luxembourg, Mongolia, the Netherlands, Nigeria, Portugal, the Russian Federation, Slovenia, Spain, Sweden, Switzerland, Ukraine, the United Kingdom, and the United States.

    Other Resources Citing This Publication