Transcript Discussion: Burda - De Nederlandsche Bank

Discussion: Burda/Hunt
Helge Berger
In a nutshell
• A very good and comprehensive paper that
takes a look at data, model, and institutions
• Miracle decomposition
40 percent = expectations
20 percent = wage moderation
40 percent = other
– private sector working time accounts
– public sector short time work
–?
Questions
•
•
•
•
•
Too narrow look at stylized facts?
Asking a lot of expectations?
Stress sectoral nature of shock enough?
Missing a longer term trend?
Coherent story including all relevant
institutions?
Only DEU vs. US?
Boom/bust countries vs. others!
18
16
Peak-to-trough declines in output and employment
14
12
10
8
Real GDP
Total employment
6
4
2
0
Note: Calculated as the percent-difference between peak and trough
during 2007Q1-2010Q3.
(Some qualifications…)
Actual vs. predicted change in unemployment
(in ppts., based on WEO, 2010, ch. 3)
Actual change
Predicted change
Germany
Norway
Japan
Italy
Switzerland
Netherlands
Belgium
Portugal
France
New Zealand
Sweden
Greece
Austria
Denmark
Canada
Finland
UK
US
Ireland
Spain
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
-1.0
Note: Based on figure 3.8 in WEO (2010). The predicted changes in
unemployment are based on Okun's law specifications that include
additional control variables, including measures of financial,
housing and equity market stress. Source: Schindler (2011).
Expectations over-stressed?
• Firms first underestimate the resilience of the
German economy (explaining ‘underhiring’
prior to the crisis)…
• …but then are firmly convinced that the “great
recession” will be quite short (explaining
‘underfiring’ during the crisis)
• Too strong assumptions?
Sectoral shock under-stressed?
• For DEU, shock was very specific
• Would expectations have mattered as much in
services sector?
• Manufacturing
– Longer planning horizon
– Larger skilled worker constraint
– Higher capacity for time accounts
Missing an improving trend…
12
11
Actual and simulated unemployment paths under different
steady-state assumptions
10
9
8
Unemployment rate (actual)
5.3%
6.0%
7.0%
8.1% (historical average)
7
6
5
Simulations*
2010Q3
2010Q1
2009Q3
2009Q1
2008Q3
2008Q1
2007Q3
2007Q1
2006Q3
2006Q1
2005Q3
2005Q1
2004Q3
2004Q1
2003Q3
2003Q1
2002Q3
2002Q1
2001Q3
2001Q1
2000Q3
2000Q1
4
Note: Hypothetical paths are model-based transition paths under the assumption
that the Hartz IV reforms lowered the structural (steady-state) unemployment
rate to various levels. See text for details. Source: Schindler (2011),
Lam (2011)
*Search model: u_t+1 = u_t (1-a) + (1-u_t) d (w/ d=1/2; u_ss; a)
…missing employment impact?
• Counterfactual would be “better” than actual
employment figures, leading to a larger
notional loss during ‘grand recession’
• Estimates:
– Pre-crisis unemployment = 7.2 percent
– Crisis peak (so far): 7.6 percent
– Simulated: about 8.2 percent
Missing institutions?
• Fiscal policy, unions,…
• Labor courts (B/Danninger 2006; B/Neugart 2011a/b)