structural policies and economic resilience toone explanation: resilience differs • resilience to...
TRANSCRIPT
Structural policies and economic resilience to shocksR. Duval, J. Elmeskov and L. Vogel (OECD)
1 March 2007
Motivation of the paper
• Business cycle fluctuations have become smaller• But it is unclear whether business cycles have become
more synchronised, except perhaps within the euro area
• Sources of cyclical divergence:- Idiosyncratic shocks- Divergent responses to common shocks: have English-
speaking and Nordic OECD countries been more “resilient” than Continental European ones lately? Is there a “double dividend” from structural reforms?
Purpose of this paper: study extent to which structural and macro policy settings affect resilience
Stylised features of business cycles
• Three simple OG measures: OECD, HP filter, Baxter-King filter
• Cross-country divergence of OGs has declined• But unclear whether this just reflects smaller amplitude
of cycles within countries or also greater synchronisation of cyclical positions:
- “Normalised” stdev of OGs has not come down- Marginally increased coincidence of turning points- No decline in idiosyncratic vs common comp. of OGs…
- …except perhaps within the euro area
])[(][ itititititGAP εγγγλεγλ +−++=++=
“common” component “idiosyncratic” component
One explanation: resilience differs• Resilience to shocks comprises at least 2 dimensions:- Extent to which shocks are damped (A)- Speed with which output reverts to potential (B)
• Labour/product market regulations (LMPMR) are expected to have ambiguous effects on resilience (e.g. cost-push or technology shocks in basic NK framework):
- Wage/price stickiness increases A but reduces B- Real wage rigidities reduce A but increase B• Strong monetary policy transmission mechanisms good for
resilience in general: they increase both A and B
This is largely an empirical issue Set up empirical framework which disentangles A and B Cross-country time-series analysis of OG patterns
Econometric methodology• Start with NLS estimation of simple panel equation:
- Coef λt is a common unobserved shock (Blanchard-Wolfers, EJ 2000) resilience to common shocks
- Coefs ϕi and are γi capture country-specific persistence and amplification mechanisms
- AR(2) specification suggested by residual auto-correlation test + sinusoidal pattern of OGs
- Sample: 20 OECD countries, 1982-2003, annual data• Alternative concepts of OG:- OECD OGs- Sensitivity analysis (needed due to bias/inconsistency
risks): HP OG, BK OG and U gap estimates; unobserved and observed shocks
( ) itiitititiit GAPGAPGAP εαγληϕ ++++−= −− 1)( 21
Persistence of shocks:1
coefficientImplied half-life of output gaps
(in years)
Amplification of shocks:1
coefficient
Estimate for the US:
USA 0.44 1.67 0.41
Estimates for other OECD countriesand test for statistical differences in coefficients with respect to the US:
AUS 0.34 1.3 0.24* *
AUT 0.47 1.8 -0.85***
BEL 0.45 1.7 -0.25*
CAN 0.27 1.1 0.34** **
CHE 0.41 1.6 -0.16*
DEU1 0.42 1.6 -0.15
DNK 0.31 1.2 -0.49* * **
ESP 0.54 2.3 -0.41** ** **
FIN1 0.49 2.0 0.24
FRA 0.50 2.0 -0.53***
GBR 0.41 1.6 -0.30*
IRL 0.49 1.9 0.37
ITA 0.47 1.9 -0.37**
JPN 0.50 2.0 -0.57***
NLD 0.50 2.0 -0.55**
NOR 0.57 2.5 -0.97*** *** ***
NZL 0.38 1.4 -0.61***
PRT 0.56 2.4 -0.38*** *** **
SWE1 0.40 1.5 0.18Time dummies yesObservations 434
R-squared 0.85Non-linear least squares.
Output gap equations with country dummies (20 OECD countries, 1982-2003)
iηϕ iγ
Preliminary estimation results• OG persistence: low in English-speaking and Nordic
countries, high in continental European countries and Japan
• Initial impact of common shocks: opposite patterns, i.e. high in English-speaking and Nordic countries, low in continental European countries and Japan
• OG persistence (initial impact of shocks) positively (negatively) correlated across countries with labour and product market regulatory indicators
• This is descriptive analysis, not econometric evidence:- In practice, coefficients are estimated over 22 observations- Assumes unchanged business cycle patterns over time
Expand econometric framework to allow for role of (time-varying) policies and institutions
Exploring the structural policy determinants of resilience
• Write persistence and amplification coefs as a function of indicators of labour/product market regulation:
where the Xj s and Xk s include:- Unemployment benefit replacement rate (UBRR)- Stringency of employment protection legislation (EPL)- Union coverage- Degree of “corporatism” in industrial relations - Stringency of product market regulation (PMR)
• Positive ϕj means Xj increases OG persistence• Positive γk means Xk increases initial impact of shock
itik
kkit
ktitit
j
jjit
jit XXGAPGAPXXGAP εαγληϕϕ ++
−++−
−+= ∑∑ −− )(1)()( ..21..
Addressing the multicollinearity issue• There is a multicollinearity issue• Here it is addressed in two alternative ways:1. “Statistical tournament”:- Estimate all possible models with 1 institution- Eliminate institution i if not significant- Based on new (reduced) set of institutions, re-estimate all
possible models with 2 institutions- Eliminate institution j if not always significant- …etc…and stop process when final model selected2. Principal component analysis construct a “synthetic”
indicator of LMPMR:Labour and product market regulation = 0.42*(replacement rate) + 0.45*(EPL) + 0.48*(union coverage) - 0.51*(low corporatism) + 0.37*(PMR)
1 2
Final model selected from statistical tournament
With synthetic indicators of labour and product market regulation
Persistence coefficients:1.085 1.065
[24.96]*** [23.82]***0.425 0.397
[11.39]*** [10.04]***
Effect of institutions on persistence:
EPL 0.128[4.77]***
PMR
Labour and product market regulation1 0.090[4.20]***
Effect of institutions on amplification of shocks:
EPL
PMR -0.537[5.93]***
Labour and product market regulation1 -0.147[2.93]***
Observations 434 434R-squared 0.83 0.83Non-linear least squares. Absolute value of t statistics in brackets.* significant at 10%; ** significant at 5%; *** significant at 1%.
Final output gap equations with labour and product market regulation indicators (20 OECD countries, 1982-2003)
kγ
φ
jϕ
η
Monetary factors and resilience• OG dynamics in aftermath of shocks depends also on
monetary policy response and strength of transmission mechanisms
• This is tentatively captured here by adding to the baseline equation with LMPMR indicator:
- Household mortgage debt: captures strength of one major transmission channel (negatively correlated with mortgage market regulation, and proxy for financial market flexibility)
- Flexible exchange rate regime dummy variable: captures whether monetary policy is able to react
- Financial intermediation (credit/stock market value traded)
Significant and robust negative effect of household mortgage debt on OG persistence
1 2 3 4 5
With synthetic indicators of labour and product
market regulation alone
'= 1 + household
mortgage debt
'= 1 + flexible exchange
rate regime
'= 1 + financial
intermediation
Final model with synthetic indicators of labour and
product market regulation and monetary factors
Persistence coefficients:1.065 1.060 1.057 0.918 1.060
[23.82]*** [22.81]*** [23.02]*** [11.76]*** [22.98]***0.397 0.392 0.395 0.382 0.392
[10.04]*** [9.54]*** [9.85]*** [7.51]*** [9.56]***
Effect of institutions on persistence:
Labour and product market regulation1 0.090 0.083 0.103 0.079 0.073[4.20]*** [3.80]*** [3.64]*** [3.71]*** [3.81]***
Household mortgage debt2 -0.546 -0.478[2.01]** [2.03]**
Flexible exchange rate regime 0.073[0.87]
Financial intermediation 0.0002[1.90]*
Effect of institutions on amplification of shocks:
Labour and product market regulation1 -0.147 -0.136 -0.195 -0.281 -0.136[2.93]*** [2.74]*** [2.91]*** [1.50] [2.77]***
Household mortgage debt2 0.076[0.13]
Flexible exchange rate regime -0.286[1.38]
Financial intermediation -0.003[1.16]
Country fixed effects yes yes yes yes yesTime dummies yes yes yes yes yesObservations 434 412 434 410 412R-squared 0.83 0.83 0.83 0.85 0.83Non-linear least squares. Absolute value of t statistics in brackets.* significant at 10%; ** significant at 5%; *** significant at 1%.
Output gap equations with synthetic indicators of labour and product market regulation(20 OECD countries, 1982-2003)
jϕ
kγ
φ
η
Back to resilience
• Estimated impact of LMPMR is ambiguous: persistence vsamplification (impact of mortgage regulation is not)
• Simulate final equation for different values of LMPMR indicator to see how the latter affects resilience according to 3 alternative criteria:
- Time needed for OG to close in aftermath of shock (A)- Cumulative output loss (B)- Output gap volatility (C)Warning: These are not welfare criteria
LMPMR increases A and B but reduces C• One can also use final equation to simulate how each
country (based on its current set of policy settings) is expected to score on A, B and C
Simulated degree of resilience (based on final equation, using 2003 values of policy and institutional indicators)
0.0
1.0
2.0
3.0
4.0
5.0
6.0
USA
CH
E
GBR NZL
CAN AU
S
DN
K
NLD JPN
DEU
NO
R
SWE
IRL
ESP
PRT
FIN
BEL
AUT
FRA
ITA
Time T needed for output to get back to potential(in years, following a 1 percentage point negative common shock to output gaps)
Simulated degree of resilience (continued)(based on final equation, using 2003 values of policy and institutional indicators)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
CH
E
NLD
DN
K
GBR USA NZL
DEU AU
S
PRT
NO
R
CAN
SWE
JPN
IRL
ESP
FIN
BEL
AUT
FRA
ITA
Cumulative output loss between 0 and T(as a percentage of output, following a 1 percentage point negative common shock to output gaps)
Simulated degree of resilience (continued)(based on final equation, using 2003 values of policy and institutional indicators)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
NLD
CH
E
DN
K
PRT
NO
R
DEU
SWE
IRL
FIN
ESP
BEL
FRA
AUT
NZL
AUS
GBR JP
N
CAN USA IT
A
Output gap volatility(number of squared standard deviations of the common shock, assuming there are no idiosyncratic shocks)
Summing up• LMPMR dampens initial impact but increases persistence• EPL (persistence) and PMR (initial impact) robust• This is theoretically consistent with view that labour/product
market rigidities primarily increase wage/price stickiness• Overall impact of LMPMR on resilience depends on criterion
used. Normative implications not explored here• Mortgage market regulation detrimental (tentative though)• Labour, product and financial market settings seem able to
account for some recent business cycle stylised facts: English-speaking/Nordics vs Continental Europeans
• Agenda for future work: - Further explore monetary transmission channels- Should one explore non-systematic effect of macro policies?
If so, is this feasible in a univariate framework?
Thank you
Tables and charts
Data before 1991 refer to Western Germany. BP gap data for Korea start only in 73:1. BP data stop in 05:4 due to the filtering method. LUX, MEX, TUR and Eastern European countries excluded.Source: OECD Economic Outlook 80 database and authors' calculations.
Cyclical convergence accross 23 OECD economies 1970-2006Standard deviation of unweighted output gaps
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
1970 75 80 85 90 95 2000 05
HP GAP BP GAP OECD GAP
Data before 1991 refer to Western Germany. BP gap data for Korea start only in 73:1. BP data stop in 05:4 due to the filtering method. LUX, MEX, TUR and Eastern European countries excluded.
Source: OECD Economic Outlook 80 database and authors' calculations.
Cyclical synchronisation accross 23 OECD economies 1970-2006Standard deviation of unweighted output gaps divided by the average absolute output gap
0.0
0.5
1.0
1.5
2.0
2.5
1970 75 80 85 90 95 2000 05
HP GAP BP GAP OECD GAP
Data before 1991 refer to Western Germany. BK gap data for Korea start only in 73:1.LUX, MEX, TUR and Eastern European countries excluded.
Source: OECD Economic Outlook 80 database and authors' calculations.
Cycles are defined as lasting five quarters at least. Turning points require that the upturn or downturn lasts at least over two subsequent quarters.
Business cycle turning points accross 23 OECD economiesAverage number of turning points over current and previous three quarters
HP output gap measure
6
5
4
3
2
1
0
1
2
3
4
5
6
1970 75 80 85 90 95 2000 05
Peaks Troughs
Data before 1991 refer to Western Germany. LUX eand SI excluded.Source: OECD Economic Outlook 80 database and authors' calculations.
Business cycle turning points accross euro membersAverage number of turning points over current and previous three quarters
HP output gap measure
Cycles are defined as lasting five quarters at least. Turning points require that the upturn or downturn lasts at least over two subsequent quarters.
3
2
1
0
1
2
3
1970 75 80 85 90 95 2000 05
Peaks Troughs
Idiosyncratic relative to common fluctuations over time
Standard deviation of idiosyncratic component relative to standard deviation of common component
HP output gap BP output gap
1973-1989 1990-2006 1973-1989 1990-2006
France 0.7 0.8 0.7 0.7 Belgium 0.9 0.7 0.9 0.6 Germany 0.7 1.1 0.6 1.0 Austria 1.0 0.9 0.9 0.9 Spain 1.0 1.0 1.0 0.9 Italy 1.1 0.8 1.0 0.8 Netherlands 1.2 0.9 0.9 0.8 Canada 1.2 1.3 1.1 1.3 United Kingdom 1.4 1.0 1.3 1.0 Japan 1.1 1.4 1.0 1.5 Denmark 1.5 1.1 1.5 0.9 Sweden 1.5 1.0 1.4 1.0 United States 1.4 1.2 1.4 1.2 Switzerland 1.6 1.0 1.6 1.0 Australia 1.4 1.4 1.4 1.4 Norway 1.6 1.4 1.6 1.3 Ireland 1.5 1.8 1.6 1.5 Portugal 2.1 1.7 2.1 1.6 Finland 1.8 2.1 1.7 2.1 Iceland 1.9 2.8 2.0 2.2 Korea 2.3 3.3 2.3 3.4 Greece 3.2 2.1 2.4 1.2 New Zealand 3.4 2.1 2.7 2.0
Average 1.5 1.4 1.4 1.3
Source: OECD Economic Outlook 80 database and calculations.
Idiosyncratic relative to common fluctuations in the euro area
Standard deviation of idiosyncratic component relative to standard deviation of common component
HP output gap BP output gap
1973-1989 1990-2006 1973-1989 1990-2006
France 0.5 0.5 0.3 0.3 Belgium 0.7 0.5 0.4 0.3 Germany 0.7 0.7 0.4 0.4 Austria 0.9 0.6 0.4 0.3 Spain 0.9 0.5 0.5 0.3 Italy 0.9 0.6 0.5 0.4 Netherlands 1.2 0.5 0.6 0.2 Ireland 1.3 1.2 0.7 0.6 Portugal 1.7 0.9 1.0 0.6 Finland 1.6 1.6 1.1 1.3 Greece 2.7 1.5 1.5 0.6
Average 1.2 0.8 0.7 0.5
Source: OECD Economic Outlook 80 database and calculations.
Persistence of shocks: coefficient Amplification of shocks: coefficient
Benefit replacement rate 0.12 -0.39*
EPL for regular contracts 0.62*** -0.43**
PMR 0.58*** -0.46**
Collective bargaining coverage 0.29 -0.23
Low corporatism -0.52*** 0.54***
Labour and product market regulation(synthetic indicator) 0.5** -0.51**
Source : Authors' estimates on the basis of country-specific persistence and amplification coefficients estimated in Table 2.1
coefficients and labour/product market policy indicatorsCross-country correlation coefficients between persistence/amplification
(based on simpleregressions of country-specific coefficients on the average value of each policy indicator over the period 1982-2003, using bootstrapped critical values to assess statistical significance)
iγiϕ
Correlation coefficients between labour and product market regulation indicators
Benefit replacement
rate
EPL PMR Collective bargaining coverage1
Low corporatism
Benefit replacement rate 1
EPL 0.29 1
PMR 0.15 0.37 1
Collective bargaining coverage1 0.52 0.44 0.40 1
Low corporatism -0.53 -0.57 -0.38 -0.48 1
1. time-invariant indicator (country average over the period 1980-2000)
Correlation coefficients, 1982-2003
1 2 3 4 5
Benefitreplacement rate
EPL Low corporatism Bargaining coverage
PMR
Persistence coefficients:1.077 1.081 1.069 1.123 1.112
[22.96]*** [24.71]*** [23.54]*** [23.98]*** [25.46]***0.395 0.425 0.390 0.393 0.399
[9.62]*** [11.19]*** [9.55]*** [9.68]*** [10.27]***
Effect of institutions on persistence:
Benefit replacement rate 0.002[0.75]
EPL 0.141[5.11]***
Low corporatism -0.279[3.40]***
Collective bargaining coverage 0.004[2.63]***
PMR 0.099[3.15]***
Effect of institutions on amplification of shocks:
Benefit replacement rate -0.007[1.09]
EPL -0.242[2.93]***
Low corporatism 0.493[2.47]**
Collective bargaining coverage -0.002[0.51]
PMR -0.508[6.15]***
Observations 434 434 434 434 434R-squared 0.82 0.83 0.82 0.82 0.83Non-linear least squares. Absolute value of t statistics in brackets.* significant at 10%; ** significant at 5%; *** significant at 1%.
Output gap equations with individual labour and product market regulation indicators (20 OECD countries, 1982-2003)
kγ
φ
jϕ
η
END