CHAPTER 2 OF THE OCTOBER 2018 WEO
Wenjie Chen, Mico Mrkaic, and Malhar Nabar,
with support from Benjamin Hilgenstock, Christopher Johns, and Yuan Zeng
THE GLOBAL ECONOMIC RECOVERY 10 YEARS AFTERTHE 2008 FINANCIAL MELTDOWN
CONTEXT: A PREVIOUS “GLOBAL” DOWNTURN (1982)
2Source: World Economic Outlook
Forty-eight economies (46 percent of global GDP) experienced output declines
THE IMMEDIATE AFTERMATH OF THE 2008 MELTDOWN
3Source: World Economic Outlook
Ninety-one economies (65 percent of global GDP) experienced output declines in 2009
MAIN QUESTIONS
Quantifying losses• Compared to pre-crisis trends, how did output evolve across countries in the aftermath of the crisis?
Channels• How did the associated components – capital, labor inputs, total factor productivity – advance after the crisis?
• Was technology adoption affected in the aftermath of the crisis?
Variation in post-crisis performance• What accounts for post-crisis variation in output losses across individual countries?
• What policies and structural attributes helped limit the damage and facilitate recovery?
4
0
0.01
0.02
0.03
0.04
0.05
-40 -20 0 20
Banking Crisis
No Banking Crisis
QUANTIFYING POST-CRISIS DEVIATIONS IN OUTPUT FROM PRE-CRISIS TRENDS
5
Sources: IMF staff calculations. Banking Crisis data: Laeven and Valencia, 2013. Note: Trend log GDP denotes extrapolated trend of potential GDP during 2000-08. Potential GDP estimated with the HP filter, lambda=100. GDP deviations from the pre-GFC trend, 2015-17.
-0.2
-0.1
0
0.1
0.2
0.3
Jan-00 Jan-04 Jan-08 Jan-12 Jan-16
HP filter
CF filter
MV filter
Log GDP
Alternative Estimates of Pre-Crisis Trends for US(2008 log GDP normalized to zero)
Post-Crisis Output Deviations from Pre-Crisis Trends by Banking Crisis Groups, 2015-17 (kernel density)
Losses appear permanent: output remains below pre-crisis trend in more than 60 percent of economies
Percent deviation from pre-crisis trend
6Sources: IMF staff calculations. Banking Crisis data: Laeven and Valencia, 2013. Note: Distribution of average deviations, 2015-17.
-0.80
-0.60
-0.40
-0.20
0.00
0.20
0.40
2000 2002 2004 2006 2008 2010 2012 2014 2016
Log investmentTrend log investment
Post-Crisis Output per Worker Deviations from Pre-Crisis Trends, 2015-17 (kernel density)
Post-Crisis Investment Deviations from Pre-Crisis TrendsMean Trajectory: Post-GFC Deviation from Real Investment Trend(2008 log investment normalized to zero)
Shortfall in other factor inputs could account for losses in labor productivity - sluggish investment
CHANGES IN LABOR INPUT CANNOT EXPLAIN OUTPUT LOSSES: SIMILAR PATTERN SEEN IN OUTPUT PER WORKER
0
0.01
0.02
0.03
0.04
0.05
-40 -20 0 20 40
Banking Crisis
No Banking Crisis
Percent deviation from pre-crisis trend
IN MANY COUNTRIES, SLUGGISH INVESTMENT POSSIBLY CONTRIBUTED TO SLOW CAPITAL ACCUMULATION…
7Sources: IMF staff calculations. Note: Distribution of average deviations, 2015-17.
Post-Crisis Capital Stock Deviations from Pre-Crisis Trends, 2015-17 (kernel density)
Capital stock shortfalls relative to pre-crisis trends: post-crisis deceleration in capital accumulation across AEs and major EMsnot just in construction sector
0
0.01
0.02
0.03
0.04
-40 -20 0 20 40
Banking Crisis
No Banking Crisis
Percent deviation from pre-crisis trend
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
Who
lesa
leRe
tail
Trad
e
Cons
truc
tion
Tran
spor
t
Man
ufac
turin
g
Agric
ultu
re
Oth
er S
ervi
ces
Min
ing
Qua
rry
Change in post- and pre-crisis growth rates in capital stock by industry,2011-14 minus 2000-07 averages (percent)
Sources: World Input-Output Database; and IMF staff calculations.
… AND SLOW TECH ADOPTION
8Sources: IMF staff calculations. Note: Distribution of average deviations, 2015-17.
Post-Crisis TFP Deviations from Pre-Crisis Trends, 2015-17 (kernel density)
TFP shortfalls relative to pre-crisis trends
Percent deviation from pre-crisis trend
0
0.01
0.02
0.03
0.04
0.05
0.06
-40 -20 0 20 40
Banking Crisis
No Banking Crisis
GROWTH ACCOUNTING SUGGESTS TFP DEVIATIONS ACCOUNT FOR LARGE SHARE OF GDP PER WORKER DEVIATIONS
9
Median share of GDP deviation accounted for by deviation in GDP per worker, 2015-17 (percent)Countries without banking crisis 70.4Banking crisis countries 80.5
TFP deviations account for a large share of GDP per worker deviationsMedian share of GDP per worker Deviation accounted for by TFP, 2015-17 (percent)Countries without banking crisis 79.3Banking crisis countries 78.2
GDP per worker deviations account for a large share of GDP deviations across all country groups
𝑦𝑦𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎
𝑦𝑦𝑝𝑝𝑝𝑝𝑝𝑝−𝐺𝐺𝐺𝐺𝐺𝐺 𝑎𝑎𝑝𝑝𝑝𝑝𝑡𝑡𝑡𝑡=
𝐴𝐴𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎
𝐴𝐴𝑝𝑝𝑝𝑝𝑝𝑝−𝐺𝐺𝐺𝐺𝐺𝐺 𝑎𝑎𝑝𝑝𝑝𝑝𝑡𝑡𝑡𝑡∗
𝑘𝑘𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎
𝑘𝑘𝑝𝑝𝑝𝑝𝑝𝑝−𝐺𝐺𝐺𝐺𝐺𝐺 𝑎𝑎𝑝𝑝𝑝𝑝𝑡𝑡𝑡𝑡
∝
DRILLING DEEPER INTO TFP:HIGHER POST-CRISIS LOSSES ASSOCIATED
WITH SLOWER INCREASES IN R&D EXPENDITURE…
10Sources: WDI, IMF staff calculations.Note: The bars depict the differences between 2014-16 average and 2011-13 average. Bar chart shows interquartile range (IQR) and lines display lesser of maximum (minimum) and +/- 1.5 times upper (lower) quartile range.
Changes in Research and Development Expenditureby Output Losses and Country Group (percent of GDP)
Countries with higher post-crisis losses – especially AEs - registered slower increases in R&D shares
Diffe
renc
es b
etw
een
2014
-16
aver
age
and
2011
-13
aver
age
0.7
0.8
0.9
1
1.1
1.2
1.3
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
High Loss Low Loss High Loss Low Loss
Advanced Economies Emerging Markets
…AND WITH SLOWER TECH ADOPTION
11
Average Change in Robot Density by Output Loss and Country Group, 2010-14
Sources: International Federation of Robotics; World Input-Output Database; and IMF staff calculations.Note: Robot density defined as robot flow / thousand hours worked. LHS bar chart shows interquartile range (IQR) and lines display lesser of maximum (minimum) and +/- 1.5 times upper (lower) quartile range.
Robot diffusion appears slower in countries with higher post-crisis output losses
-0.02
-0.01
0
0.01
0.02
0.03
0.04
High Loss Low Loss High Loss Low Loss
Advanced Economies Emerging Markets
+ +
Differences in medians between high/low output loss samples among AEs statistically significant at 10 percent.
+
CROSS-COUNTRY VARIATION IN POST-CRISIS PERFORMANCE: DEVIATIONS ARE PERSISTENT
12
• Deviations are persistent over time
• The correlations between GDP deviations for 2011-13 and 2015-17 are around 0.90 (0.84 for the Spearman rank correlation).
Correlation of GDP Deviations Between Periods
Source: IMF staff calculations.
2011-13
2015
-17
-80
-60
-40
-20
0
20
40
-30 -20 -10 0 10 20 30
CORRELATES OF CROSS-COUNTRY VARIATION IN POST-CRISIS PERFORMANCE
13
• Building on WEO 2009; Lane and Milesi-Ferretti 2010, 2014; Claessens, Dell’Ariccia, Igan, and Laeven 2010; Gourinchas and Obstfeld 2012; Cerra, Panizza, and Saxena 2013…
• OLS specification; 120 economies• ∆𝑦𝑦𝑖𝑖 : output deviations 2011-13 and 2015-17• 𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄 : initial conditions averaged over 2005-2008
– Macrofinancial vulnerabilities– Flexibility to adjust and economic structure– Initial policy space– Banking crisis
• Post-Crisis Policy Actions 2008-2009– Capital Injection– Guarantees– Total stimulus
∆𝑦𝑦𝑖𝑖= 𝛼𝛼 + 𝜷𝜷 ∗ 𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝒄𝑖𝑖 + 𝜀𝜀𝑖𝑖
CORRELATES OF CROSS-COUNTRY VARIATION IN POST-CRISIS PERFORMANCE
14
Domestic Credit Growth – ** – *** – *** – *** – *** – **Demand Exposure to Advanced Economies – *** – + + – –Demand Exposure to China + + + + * + ** +Financial Openness – * – – – – –CA Balance + + *** –CA Gap + *** + *** +Share of Manufacturing in GDP + + +Difficulty of Dismissal – ** – * – **Precrisis GG Debt Change – *** – *** – ***De Facto Peg Dummy – ** – *** –Banking Crisis – ** –Source: IMF staff calculations.
Table 2.2. Impact of Precrisis Conditions on 2011–13 GDP Deviations from Precrisis Trend
All Countries AEs EMs
Note: + denotes positive impact, – denotes negative impact. Precrisis conditions are averaged over 2005–08. Results in columns (1) and (2) are reported in Online Annex Table 2.2.5. Results in columns (3) through (6) are reported in Online Annex Table 2.2.7. AEs = advanced economies; CA = current account; CA Gap = excess external balance, Lee and others (2008); EMs = emerging markets; GG = general government.*** p < 0.01, ** p < 0.05, * p < 0.1.
(1) (2) (5)(3) (4) (6)
STRONGER PRE-CRISIS BANKING REGULATION: LOWER PROBABILITY OF BANKING CRISIS
15
Probability of banking crisis in 2007-09 is lower in economies with stronger pre-crisis banking regulation
Banking Regulation Index: Barth, Caprio, Levine 2013
Source: IMF staff calculations.
Prob
abilit
y of
Ban
king
Cris
is
Strength of Banking Regulation
0
0.2
0.4
0.6
0.8
-2 -1 0 1 2
Probit
Logit
Linear Probability Model
Stronger banking regulation
CORRELATES OF POLICY INTERVENTIONS AND CROSS-COUNTRY VARIATION IN POST-CRISIS PERFORMANCE
16
Impact on 2015-17 GDP deviations from one standard deviation increase in drivers
Percent deviation from pre-crisis trend
Source: IMF staff calculations.
Less negative / more positive deviation from trend
Notes:*** p<0.01, ** p<0.05, * p<0.1.
0 1 2 3 4
Upfront Government Financing
Guarantees(Excl. Deposit Insurance)
Purchase of Assets andLending by Treasury
Capital Injections
Total Headline Support forFinancial and Other Sectors
*
***
INEQUALITY INCREASED MORE IN COUNTRIES THAT EXPERIENCED LARGER OUTPUT LOSSES
17Sources: Standardized World Income Inequality Database (Solt 2016); and IMF staff calculations.
-8
-6
-4
-2
0
2
4
6
8
-30 -20 -10 0 10 20 30
Aver
age
chan
ge in
Gin
i coe
fficie
nt
Percent deviations from precrisis trend
Change in Inequality (2014-15 relative to 2005-08) and Postcrisis Output Deviations (2011-13)
LOWER FERTILITY RATES IN SOME COUNTRIES, WITH IMPLICATIONS FOR FUTURE LABOR INPUT
18
1
1.2
1.4
1.6
1.8
2
2.2
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
OECD AE EM US Italy Spain Greece
Total Fertility Rate(Number of births per woman)
Sources: Organisation for Economic Co-operation and Development (OECD); World Bank, World Development Indicators database; and IMF staff calculations. Note: OECD is the average fertility rate for OECD and partner countries. AEs = OECD and partner advanced economies; EMs = OECD and partner emerging market economies. See Online Annex 2.1 for country list.
CRISIS AFTERMATH: TAKEAWAYS
19
• Persistence of losses following the crisis – widespread, not just in countries with banking crisis
• Sluggish investment is a key channel – associated with long-lasting capital and TFP shortfalls
• R&D investment increased less and tech adoption appears slower in countries that suffered larger losses
POLICY IMPLICATIONS
20
MACROFINANCIAL AND EXTERNAL
• More rapid pre-crisis credit growth associated with larger post-crisis losses
• Stronger external balances associated with lower post-crisis losses
• Stricter banking regulation associated with lower probability of banking crisis
FISCAL, MONETARY, STRUCTURAL
• Fiscal buffers help reduce GDP damages
• Less rigid exchange rate regimes help lessen GDP damages
• Labor market rigidity can slow the pace of recovery; associated with larger displacement effects of automation
POST-CRISIS ACTIONS
• Capital injections mitigate post-crisis GDP loss
• Guarantees help lessen GDP damages