fabrizio balassone and maura francese (*)

12
Fabrizio Balassone and Maura Francese (*) Cyclical Asymmetry in Fiscal Policy, Debt Accumulation and the Treaty of Maastricht (*) Banca d’Italia - Research Department

Upload: giles

Post on 22-Jan-2016

22 views

Category:

Documents


0 download

DESCRIPTION

Cyclical Asymmetry in Fiscal Policy, Debt Accumulation and the Treaty of Maastricht. Fabrizio Balassone and Maura Francese (*). (*) Banca d’Italia - Research Department. Structure of the presentation 1 - Motivation 2 - The stylised framework 3 - Empirical results - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Fabrizio Balassone and Maura Francese (*)

Fabrizio Balassone and Maura Francese (*)

Cyclical Asymmetry in Fiscal Policy, Debt Accumulation and the Treaty of Maastricht

(*) Banca d’Italia - Research Department

Page 2: Fabrizio Balassone and Maura Francese (*)

Structure of the presentation

1 - Motivation

2 - The stylised framework

3 - Empirical results

3.1 - Asymmetry in fiscal policy

3.2 - Impact of EMU fiscal rules

3.3 - Debt accumulation

4 - Conclusion and Possible extensions

Page 3: Fabrizio Balassone and Maura Francese (*)

Motivation

three issues on which evidence is scarce and not conclusive

a - is fiscal policy asymmetric over the cycle?

EC (2001), Buti et al. (1998), Buti-Sapir (1998),

b - if so, how did this contribute to debt accumulation?

Melitz (2002)

c -did EMU fiscal rules affect fiscal reaction to the cycle?

Galì-Perotti (2003)

Page 4: Fabrizio Balassone and Maura Francese (*)

Stylised framework: (1) fiscal balance components

1) bt = blt + bc

t

bt budget balance (with bt>0 indicating a deficit)bl

t long-run component bc

t cyclical component

2) blt = bt-1 + ( b* - bt-1) + (d* - dt-1) ,>0

dt government debtb*, d* preferred levels; d*=b*/kk trend nominal growth

3) bct = E[t]

t output gap

Page 5: Fabrizio Balassone and Maura Francese (*)

Stylised framework: (2) asymmetry in fiscal policy

4) bct = p E[t

P] + n E[tN]

pn

E[t] if E[t] >0E[t

P] = 0 if E[t] <0

E[t] if E[t] <0E[t

N]= 0 if E[t] >0

Page 6: Fabrizio Balassone and Maura Francese (*)

Stylised framework: (3) the estimating equation

5) bt= 0+1dt-1+ 2bt-1+p E[tP] + n E[t

N]

where 0=(+/k)b* 1=- 2=(1-)

Stabilisation policy n,p<0

Index of asymmetry in the conduct of fiscal policy:

6) = n - p

Page 7: Fabrizio Balassone and Maura Francese (*)

The data

sample: EU countries (excluding Luxembourg), USA and JP

coverage for b and d ranges from 1969-2002 to 1977-2002

output gap: E[t] = t

t: HP filter applied to GDP (1960-2004) drop from 2001 (end point bias)

Page 8: Fabrizio Balassone and Maura Francese (*)

Empirical results: (1) Asymmetry in fiscal policy

Table 1 - Full sample

Variable Fixed effect model - OLS

Arellano-Bond estimation

Constant 1,310 -0,006

(0,246) (0,013)Lagged Debt (dt-1) -0,022 -0,020

(0,004) (0,006)Lagged Balance (bt-1) 0,884 0,883

(0,027) (0,026)

Current Positive Cycle t mt -0,052 -0,054

(0,078) (0,074)

Current Negative Cycle t (1-mt) -0,416 -0,414

(0,080) (0,075)

Observations 466 450

Ratcheting n

p -0,364 -0,360

test: =0 (p-value in brackets) -2,69 (0,007) 8,01 (0,005)

b* with k=0,05 2,396

Page 9: Fabrizio Balassone and Maura Francese (*)

Table 2 - EU countries (excluding Luxembourg)

Variable Arellano-Bond Fixed effect

OLS

Constant 0,026 1,045

(0,016) (0,271)Lagged Debt (dt-1) before Maastricht -0,013 -0,008

(0,006) (0,006)Lagged Debt (dt-1) after Maastricht -0,029 -0,022

(0,006) (0,004)Lagged Balance (bt-1) 0,841 0,844

(0,028) (0,030)

Current Positive Cycle t mt -0,129 -0,118

(0,791) (0,085)

Current Negative Cycle t (1-mt) -0,416 -0,408

(0,079) (0,085)

Observations 391 405

Ratcheting n

p -0,287 -0,290

test: =0 (p-value in brackets) 4,58 (0,032) -2,00 (0,046)b*=

0/[(1- (1/k)] with k=0,05 - before Maastricht 3,212b*=

0/[(1- (1/k)] with k=0,05 - after Maastricht 1,776

test: 1a92-1b92 =0 (p-value in brackets) 21,54 (0,000) 4,10 (0,000)

Empirical results: (2) Impact of EMU rules

Page 10: Fabrizio Balassone and Maura Francese (*)

Empirical results: (3) Debt Accumulation

(7) dt= dt-1/(1+kt) + b^t + st where b^t = OLS fitted valuesst = stock-flow adj.

Table 3 - Actual and simulated debt accumulation Av 16 Av EU DE IT PT USA

Initial government debt a 34,51 33,73 18,57 37,92 15,32 49,30

Actual Government debt in 2000 b 68,12 64,75 60,19 110,62 53,27 59,88

Observed variation c=b-a 33,61 31,02 41,62 72,69 37,95 10,58

Simulated debt in 2000: Asymmetric fiscal policy d 61,59 59,09 51,16 105,59 55,71 59,78

Symmetric fiscal policy e 51,75 49,16 41,88 98,90 41,85 51,93

Induced Debt Accumulation f=d-e 9,84 9,93 9,28 6,69 13,86 7,85

% of observed variation 29,3 32,0 22,3 9,2 36,5 74,2

b* before 1992 3,21 1,93 6,70 3,86b* after 1992 1,78 1,07 3,70 2,13

2,40 2,47

Page 11: Fabrizio Balassone and Maura Francese (*)

Debt Accumulation: a graphical illustration

Italy - Debt GDP ratio

30

40

50

60

70

80

90

100

110

120

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

asymmetric symmetrict

dt

no reaction

asymmetric

symmetric

dt* a

dt* s

1 2

Page 12: Fabrizio Balassone and Maura Francese (*)

Conclusions and possible extensions

Main findings:

- significant cyclical asymmetry in fiscal policy (n=-0.4; p=0.0)

- sizeable contribution to debt accumulation (about one third)

- no effect from EMU fiscal rules

Possible extensions:

- separate analysis of revenue and expenditure

- distinction between automatic and discretionary reaction

- robustness to different measures of expected output gap

- specification of control variables to account for different governments and institutional settings