Upload
aurelien-poissonnier
View
31
Download
0
Embed Size (px)
Citation preview
Soutenance de thèse
Essays on Macroeconomics
Aurélien Poissonnier([email protected])
Directeur de Thèse : Édouard Challe
École Polytechnique-Insee-Crest
4 décembre 2015
A. Poissonnier (Polytechnique) Soutenance Dec.4 1 / 35
Chapter 1
Consumption, interest rates and
expectations:
a reconciliation
A. Poissonnier (Polytechnique) Soutenance Dec.4 2 / 35
Motivation
A challenge for monetary policy models (Canzoneri et al., 2007)
Euler equation equalizes the monetary policy rate with themarginal rate of intertemporal substitution of consumption (MRS)
US data reject this equality (null or negative correlation)
A. Poissonnier (Polytechnique) Soutenance Dec.4 3 / 35
Comparison of Fedfunds and MRS
1970 1975 1980 1985 1990 1995 2000 2005 2010
0
1
2
3
4
MRSFedfunds
Baseline �t to the data: R2adjusted = −1%, no signi�cant coe�cient
A. Poissonnier (Polytechnique) Soutenance Dec.4 4 / 35
Methodology
Full model = Euler equation + expectations + choice of interest rateThe rejection is conditional on
1 households react to monetary policy rate (baseline model)2 expectations = VARX forecast
New tests
1 households may react to another interest rate⇒ each household may arbitrate on a speci�c market (limitedparticipation)
2 baseline method is bias towards rejecting (Euler)⇒ joint MLE (Euler+VARX)
A. Poissonnier (Polytechnique) Soutenance Dec.4 5 / 35
Households speci�c rates
Fedfunds only or with mortgage rate + car loans rate + personal loansrate + 3 month treasury bill (+ deposit rate + return on money)
1 rate 5 ratesFedfunds
log-lik. 1947.81 1977.37R2adj (Euler) -1 32
Estimation yields a sensible mix of interest rates:20% treasury bill + 80% personal loans
A. Poissonnier (Polytechnique) Soutenance Dec.4 6 / 35
Optimal (Rational) expectations ?
Not GMM but MLE → possible to1 depart from rational expectations paradigm
joint estimation improves the �t: R2adj = 46%
2 characterize expectationssmoothed MRS = smoothed expectations
A. Poissonnier (Polytechnique) Soutenance Dec.4 7 / 35
Joint estimation
1975 1980 1985 1990 1995 2000 2005 2010
0
1
2
3
4
FedfundsPers. LoansMRSOptimal rate combination
Figure: Interest rates and MRS from a joint estimation
A. Poissonnier (Polytechnique) Soutenance Dec.4 8 / 35
Intuitions from joint estimations
→ behavioural equation: measuring expectations with survey dataimplies more co-�uctuations→ structural break: after the great moderation, di�erent expectationprocess→ inattentive consumers: when in�ation is steadily low, it can beoptimal to form naive expectations
A. Poissonnier (Polytechnique) Soutenance Dec.4 9 / 35
Yes, the Euler equation can be used to describe householdsconsumption in a monetary policy model but
US households respond to the interest rate on personal
loans (lower savings, widespread use of credit cards)
French households respond to the interest rate on regulatedsavings (90% pop. holds a livret)
⇒ weaker transmission of monetary policy
Augmented model
⇒ inattentiveness: forecasting consumption is relatively costless,but in�ation is more complex. When the latter is steadily low, itcan be optimal to form naive expectations.
+ structural break around 1990
A. Poissonnier (Polytechnique) Soutenance Dec.4 10 / 35
A reconciliation
0
1
2
3
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012
Euler equation res.MRS1MRS2Optimal rateFedfund
Figure: Interest rates and MRS (inattentive consumer + treasury bill &personal loans & survey + precautionary savings)
A. Poissonnier (Polytechnique) Soutenance Dec.4 11 / 35
Chapter 2
The Taylor principle is valid under wage
stickiness
A. Poissonnier (Polytechnique) Soutenance Dec.4 12 / 35
Outline
Motivation
textbook model of monetary policy with staggered prices andwages
a policy rule reacting to wage in�ation is welfare improving (Galí,2008, chapter 6; Erceg et al., 2000)
does the Taylor principle hold in this set-up ?
Result the frontier of determinacy of the model is
Φp + Φw + Φy(1− β)(λp + λw)
κwλp + κpλw> 1 (1)
A. Poissonnier (Polytechnique) Soutenance Dec.4 13 / 35
Interpretation
Frontier symmetric in price and wage in�ation
generalize the frontier of determinacy with staggered prices only(Woodford, 2001)
generalize the frontier of determinacy in continuous time (Flaschelet al., 2008)
the Central Banker should react more than one for one to
permanent changes in in�ation (Woodford, 2011, chapter 4)
the Central Banker should/could not just focus on price in�ation
A. Poissonnier (Polytechnique) Soutenance Dec.4 14 / 35
Chapter 3
Households Satellite Account for France
Methodological issues on the assessment
of domestic production
with D. Roy
A. Poissonnier (Polytechnique) Soutenance Dec.4 15 / 35
Motivation
System of National Accounts does not record all economic activity
Old debate on the frontier of production (and more fundamentallyon the purpose of SNA: economics or welfare?) (Vanoli, 2002)
Speci�c literature on the valuation of domestic work
Recent focus on Households Satellite Accounts (Eurostat, 2003;Stiglitz et al., 2009) but methodological issues remain
This chapter illustrates such methodological issues on the example ofFrance in 1998 and 2010.
A. Poissonnier (Polytechnique) Soutenance Dec.4 16 / 35
Revision of macroeconomic �gures
Macroeconomic levels are scaled-up
GDP: +33% (+31% in 1998)
disposable income +50% (+49% in 1998)
consumption +58% (+56% in 1998)
savings ratio -4 pp (idem in 1998)
Growth is scaled down
GDP: +15% in HHSA, +20% in SNA
purchasing power of gross disposable income: +17% in HHSA,+27% in SNA
A. Poissonnier (Polytechnique) Soutenance Dec.4 17 / 35
Market or non-market?
Cooking 459 bn e - Restaurants 59 bn e
Cleaning 253 bn e - Housekeepers 6 bn e
Laundry 61 bn e - Dry cleaners 2 bn e
Driving 38 bn e - Transport 27 bn e
Care 92 bn e - Social Services 67 bn e
A. Poissonnier (Polytechnique) Soutenance Dec.4 18 / 35
Methodological issues, two examplesProductive leisures
Overall at least as many hours of domestic work as paid working timeBut the de�nition can double this time
Table: Average working time per person for three possible perimeters ofdomestic work
Perimeter core (I) intermediate (II) extensive (III)
Daily 2 h 07 3 h 04 3 h 53Weekly 14 h 50 21 h 30 27 h 14
Share ofvolunteerwork
3.7% 3.8% 5%
Women'sShare
72% 64% 60%
Coverage: individuals aged 11 and over, France (excl. French Guyana and Mayotte).Source: Insee, 2010 Time Use Survey.
A. Poissonnier (Polytechnique) Soutenance Dec.4 19 / 35
Methodological issues, two examplesImputed taxes and the savings ratio
In 2010 SNA, the savings ratio = 15.9%
no taxes on domestic production and associated income = 13.2%
domestic production includes taxes but taxes are not paid onassociated income = 11.5%
taxes are paid on income from domestic production = -1.6%
A. Poissonnier (Polytechnique) Soutenance Dec.4 20 / 35
Chapter 4
Structural reforms in DSGE models:
a plead for sensitivity analyses
with B. Campagne
A. Poissonnier (Polytechnique) Soutenance Dec.4 21 / 35
Motivation
Increasing use of DSGEs by governments, central banks andinternational institutions to assess the impact of structural reforms:
Quest III (EC), NAWM or EAGLE (ECB), GEM (IMF), SIGMA(FED)...
However, what can these business cycle models tell us aboutstructural reforms?
A. Poissonnier (Polytechnique) Soutenance Dec.4 22 / 35
Our main results
Smets and Wouters' model in line with Blanchard and Giavazzi (2003)
Consumption-leisure arbitrage plays a key role in the transmission ofstructural reforms to the economy
⇒ Frisch elasticity (+ risk aversion and habits) are quantitatively crucial
Welfare analysis shows households might loose upon reform
Need for sensitivity analysis
More fundamental rethink of labour market modelling and the di�erencesbetween short term and long term elasticities.
A. Poissonnier (Polytechnique) Soutenance Dec.4 23 / 35
Channels of transmission of structural reforms...in Smets and Wouters' model
K
¯RWrk
L
Capital market Labour market
DD
Capital demand curve
θ−1θ αζ1−α
(KL
)α−1
= rk
Capital supply curve
rk = 1+rΠ− 1 + δ
Labour demand curve
θ−1θ (1− α)ζ1−α
(KL
)α= RW
Labour supply curve
RW = θwθw−1 (1 + σl)cf(l)
A. Poissonnier (Polytechnique) Soutenance Dec.4 24 / 35
Channels of transmission of structural reforms...in Smets and Wouters' model
K
¯RWrk
L
Capital market Labour market
D
S
D
Capital demand curve
θ−1θ αζ1−α
(KL
)α−1
= rk
Capital supply curve
rk = 1+rΠ− 1 + δ
Labour demand curve
θ−1θ (1− α)ζ1−α
(KL
)α= RW
Labour supply curve
RW = θwθw−1 (1 + σl)cf(l)
A. Poissonnier (Polytechnique) Soutenance Dec.4 24 / 35
Channels of transmission of structural reforms...in Smets and Wouters' model
K
¯RWrk
L
Capital market Labour market
D
S
S
D
Capital demand curve
θ−1θ αζ1−α
(KL
)α−1
= rk
Capital supply curve
rk = 1+rΠ− 1 + δ
Labour demand curve
θ−1θ (1− α)ζ1−α
(KL
)α= RW
Labour supply curve
RW = θwθw−1 (1 + σl)cf(l)
A. Poissonnier (Polytechnique) Soutenance Dec.4 24 / 35
Channels of transmission of structural reformsProduct market deregulation : ↑ substitutability between goods
dividends ↓ and factors' remuneration ↑ (D1 to D')
increased demand for factors translates into ↑ marginal productivity andfurther ↑ demand (D' to D2)
positive wealth e�ect, consumption ↑ and labour supply ↓ (S1 to S2)
K
¯RWrk
L
Capital market Labour market
D1
D'
D2
S2S1
S1
D1
D'D2
At steady state
capital ↑,labour ↑,production ↑
consumption↑, real wages ↑
A. Poissonnier (Polytechnique) Soutenance Dec.4 25 / 35
Channels of transmission of structural reformsLabour market deregulation : ↑ substitutability between workers
lower marker power of workers implies an increased labour supply for asame wage (S1 to S')
increased labour supply translates into ↑ marginal productivity of capitaland then labour (D1 to D2)
consumption-leisure arbitrage adjusts (S' to S2 - wealth e�ect) but
unchanged real wages in the long-run
At steady state
capital ↑, labour↑, production ↑ inthe sameproportions
consumption ↑,real wagesstagnates
K
¯RWrk
L
Capital market Labour market
D1
D2
S1S2
S'S1
D1
D2
A. Poissonnier (Polytechnique) Soutenance Dec.4 26 / 35
Channels of transmission of structural reformsQuantitative simulations
Figure: Production Figure: Real wage Figure: Total utility
Steady state variations upon reforms in p.p.
Best practices = standard exercise of convergence towards the threebest EU performers (UK, SW, DK)
A. Poissonnier (Polytechnique) Soutenance Dec.4 27 / 35
Results
Long term output gains but transitory costs
Output enhancing but ambiguous sign for utility
Importance of the households' utility and habits calibration
Additive or multiplicative habits identical in short run butequivalent to change in Frisch and elasticity of substitution in longrun
A. Poissonnier (Polytechnique) Soutenance Dec.4 28 / 35
Conclusion
Four chapters
households behaviour
monetary policy
DSGE modelling
normative and descriptive results
link with the data
A. Poissonnier (Polytechnique) Soutenance Dec.4 29 / 35
Appendix
A. Poissonnier (Polytechnique) Soutenance Dec.4 30 / 35
Test of the Euler equation by maximum likelihood
ifedt − Etπt+1 = δ + σEt∆ct+1 +∑x 6=fed
γx(ixt − i
fedt
)+ ζt (Euler)
Yt+1 = [ct+1, πt+1] = EtYt+1 + εt+1 = Λ(L)Yt + ΓXt + εt+1 (VARX)
Baseline
Hyp. expectation errors =VARX residuals
Hyp. households react toFedfunds (γx = 0)
Method �rst MLE (VARX)then MLE (Euler)
New tests
households may react to adi�erent interest rate⇒ γx 6= 0
Baseline method is biastowards rejecting (Euler)⇒ joint MLE(Euler+VARX: [ε, ζ])
A. Poissonnier (Polytechnique) Soutenance Dec.4 31 / 35
A monetary policy model with staggered prices and wages
πpt = βE(πpt+1|t) + κpyt + λpωt (2)
πwt = βE(πwt+1|t) + κwyt − λwωt (3)
ωt−1 = ωt − πwt + πpt + ∆ωt (4)
yt = E(yt+1|t)−1
σ(it − E(πpt+1|t)− r
nt ) (5)
it = Φpπpt + Φwπ
wt + Φyyt + vt (6)
A. Poissonnier (Polytechnique) Soutenance Dec.4 32 / 35
λp =(1−θp)(1−βθp)
θp1−α
1−α+αεp , withI 0 < θp < 1, is the Calvo parameter on pricesI 0 < α < 1, is the Cobb-Douglas parameterI 0 < εp ≤ 1, is the mark-up on goods
⇒ 0 < λp
λw = (1−θw)(1−βθw)θw(1+ϕεw) , with
I 0 < θw < 1, is the Calvo parameter on wagesI 0 < ϕ, is the Frisch elasticityI 0 < εw ≤ 1, is the mark-up on wage setting
⇒ 0 < λw
κp =αλp1−α
κw = λw(σ + ϕ1−α)
A. Poissonnier (Polytechnique) Soutenance Dec.4 33 / 35
Channels of transmission of structural reformsWhat should we expect?
Blanchard and Giavazzi (2003)
I monopolistic competition, entry costs and wage Nash bargainingI production = employment
Reform short-run transition long-run
↓ bargaining power of workerspro�ts ↑, and real
wages ↓
more �rms enter,labour and real
wages ↑
labour ↑ and realwages stagnate
↑ substitutability between goodsreal wages and
labour ↑pro�ts ↓ so �rms
exitneutral
↓ entry costsneutral as �xednumber of �rms
�rms start enteringreal wages and
labour ↑
A. Poissonnier (Polytechnique) Soutenance Dec.4 34 / 35
Long term elasticities of output and utility
Elasticities when performing :
a product market deregulationdYY
=(
α1−α + 1
(1+σl)(1−hl)(1+B) (1 + iycy ))
dθθ
θ−1 > 0
dUU
= (1− σc)(1− hc)[(
α1−α −
iycy
)+
1− θw−1θw
θ−1θ
(1−α)(1−hl)(1−hc)cy
(1+σl)(1−hl)(1+B) (1 + iycy )
]dθθ
θ−1
a labour market deregulationdYY
= 1(1+σl)(1−hl)(1+B)
dθwθw
θw−1 > 0
dUU
= (1− σc)(1− hc)1− θw−1
θw
θ−1θ
(1−α)(1−hl)(1−hc)cy
(1+σl)(1−hl)(1+B)
dθwθw
θw−1
A. Poissonnier (Polytechnique) Soutenance Dec.4 35 / 35