Fiscal Sustainability Hypothesis Test in Central and Eastern Europe: A Panel Data Perspective

Owusu, Benjamin

doi:10.2478/ceej-2021-0021

Figures & Tables

Wald test of coefficient and confidence intervals

Models	MG-FMOLS	MG-DOLS
T stat	−32.018^***	−49.89^***
Chi-square (1 df)	1,025.13	1,754.72
P-value	0.000	0.000
95% confidence intervals	(0.934–0.941)	(0.932–0.938)

Panel summary Statistics

	Revenue-GDP ratio	Expenditure-GDP ratio	Debt-GDP ratio
Mean	0.391	0.419	0.340
Standard Deviation	0.039	0.505	0.178
Maximum	0.482	0.541	0.711
Minimum	0.308	0.321	0.038
Observations	250	250	250

Deficit to GDP ratio

Year	Czh	Est	Hun	Lat	Lith	Pol	Svk	Slvn	Bulg	Rom
1995	−12.44	1.05	−8.60	−1.43	−1.53	−4.26	−3.47	−8.15	−5.52	−2.00
1996	−3.01	−0.32	−4.38	−0.42	−3.22	−4.63	−9.82	−1.09	−8.11	−3.57
1997	−3.19	2.15	−5.55	1.42	−11.59	−4.61	−6.27	−2.31	0.76	−4.43
1998	−4.19	−0.73	−7.39	0.03	−3.03	−4.21	−5.30	−2.33	1.08	−3.24
1999	−3.14	−3.29	−5.23	−3.74	−2.82	−2.28	−7.17	−2.97	0.09	−4.42
2000	−3.57	−0.04	−2.980	−2.73	−3.17	−2.98	−12.63	−3.65	−0.53	−4.60
2001	−5.48	0.20	−3.94	−1.95	−3.51	−4.77	−7.22	−4.45	1.05	−3.46
2002	−6.36	0.42	−8.76	−2.29	−1.85	−4.85	−8.22	−2.37	−1.16	−1.93
2003	−6.89	1.82	−7.11	−1.46	−1.27	−6.08	−3.12	−2.56	−0.39	−1.43
2004	−2.39	2.34	−6.52	−0.92	−1.39	−5.04	−2.32	−1.94	1.80	−1.09
2005	−2.9	1.08	−7.72	−0.36	−0.34	−3.96	−2.87	−1.32	1.00	−0.81
2006	−2.17	2.87	−9.21	−0.49	−0.27	−3.56	−3.58	−1.23	1.81	−2.14
2007	−0.65	2.73	−5.03	−0.51	−0.81	−1.85	−2.05	−0.05	1.10	−2.73
2008	−1.98	−2.65	−3.73	−4.20	−3.09	−3.60	−2.52	−1.39	1.59	−5.35
2009	−5.45	−2.16	−4.69	−9.49	−9.13	−7.25	−8.15	−5.81	−4.05	−9.06
2010	−4.19	0.19	−4.39	−8.60	−6.92	−7.40	−7.46	−5.60	−3.13	−6.92
2011	−2.73	1.06	−5.19	−4.25	−8.95	−4.88	−4.46	−6.63	−1.98	−5.43
2012	−3.93	−0.29	−2.27	−1.22	−3.15	−3.74	−4.37	−4.0	−0.32	−3.65
2013	−1.25	0.18	−2.54	−1.17	−2.61	−4.18	−2.87	−14.58	−0.43	−2.10
2014	−2.10	0.70	−2.76	−1.44	−0.62	−3.65	−3.11	−5.51	−5.43	−1.19
2015	−0.61	0.14	−1.97	−1.36	−0.27	−2.62	−2.67	−2.85	−1.72	−0.61
2016	0.72	−0.52	−1.76	0.06	0.23	−2.37	−2.48	−1.94	0.09	−2.62
2017	1.56	−0.77	−2.38	−0.52	0.45	−1.46	−0.95	−0.01	1.10	−2.64
2018	1.09	−0.56	−2.29	−0.74	0.60	−0.24	−1.06	0.77	1.75	−2.96
2019	0.75	−0.30	−1.83	0.51	0.13	−1.16	−1.03	0.72	−0.11	−3.33

Panel Stationarity stationarity test with sharp breaks - – Robustness robustness check

Panel - – Panel test (revenue)
Model	Level	First difference
Break (Homogeneous)	1.762 (0.039)^**	0.426 (0.335)
Breaks (Heterogeneous)	2.975 (0.001)^***	0.152 (0.440)
Panel B – Panel stationarity test (expenditure)
Model	Level	First difference
Break (Homogeneous)	1.687 (0.046)^**	0.137 (0.446)
Breaks (Heterogeneous)	2.031 (0.021)^**	0.630 (0.264)

Panel stationarity test – Expenditure

Countries	Constant			Constant and trend
Countries	k = 1	k = 2	k = 3	k = 1	k = 2	k = 3
Czechia	0.070	0.309	0.302	0.050^*	0.059	0.050
Estonia	0.099	0.410^*	0.331	0.052^*	0.124^*	0.125^*
Hungary.	0.203^**	0.298	0.201	0.044	0.101	0.091
Latvia	0.183^**	0.504^**	0.414^*	0.051^*	0.080	0.083
Lithuania	0.052	0.055	0.116	0.051^*	0.043	0.051
Poland	0.182^**	0.496^*	0.416^*	0.053^*	0.056	0.076
Slovakia	0.048	0.219	0.197	0.040	0.145^**	0.139^*
Slovenia	0.059	0.189	0.322	0.054^*	0.097	0.085
Bulgaria	0.148^**	0.089	0.095	0.051^*	0.088	0.087
Romania	0.124	0.313	0.181	0.053^*	0.136^**	0.133^*
Panel statistic	2.995^***	3.508^***	2.282^***	4.938^***	3.309^***	2.460^***
	(0.001)	(0.000)	(0.011)	(0.000)	(0.000)	(0.007)

Panel stationarity Test test -– First first difference of Revenuerevenue

Countrs	Constant			Constant and trend
Countrs	k = 1	k = 2	k = 3	k = 1	k = 2	k = 3
Czechia	0.127	0.075	0.074	0.044	0.057	0.074
Estonia	0.093	0.082	0.117	0.051^*	0.071	0.111
Hungary.	0.038	0.151	0.159	0.031	0.148^**	0.153^**
Latvia	0.063	0.062	0.107	0.047	0.052	0.056
Lithuania	0.081	0.122	0.079	0.061^**	0.082	0.077
Poland	0.088	0.128	0.157	0.057^**	0.056	0.060
Slovakia	0.108	0.531^**	0.423^*	0.076^***	0.120^*	0.099
Slovenia	0.061	0.108	0.117	0.057^**	0.101	0.077
Bulgaria	0.180^**	0.327	0.366^*	0.027	0.149^**	0.159^**
Romania	0.038	0.065	0.083	0.035	0.048	0.041
Panel statistic	1.282	0.574	0.292	4.723^***	2.965^***	2.357^***
	(0.100)	(0.283)	(0.385)	(0.000)	(0.002)	(0.009)

Chow test of poolability of coefficients (5% significance level)

Restricted model	Intercept	F-statistics	P-value	Verdict
FE (within)	Variable	2.8438	0.004	Not poolable
Pooled OLS	Fixed	14.892	0.000	Not poolable

Panel stationarity test – first difference of expenditure

Countries	Constant			Constant and trend
Countries	k = 1	k = 2	k = 3	k = 1	k = 2	k = 3
Czechia	0.195^**	0.170	0.132	0.059^**	0.094	0.093
Estonia	0.032	0.033	0.065	0.031	0.035	0.031
Hungary.	0.078	0.091	0.110	0.048^*	0.089	0.107
Latvia	0.057	0.126	0.070	0.046	0.050	0.046
Lithuania	0.078	0.148	0.080	0.053^*	0.097	0.060
Poland	0.075	0.044	0.055	0.045	0.038	0.051
Slovakia	0.122	0.136	0.137	0.047	0.083	0.088
Slovenia	0.064	0.131	0.097	0.060^**	0.111^*	0.096
Bulgaria	0.069	0.114	0.123	0.031	0.091	0.096
Romania	0.053	0.069	0.181	0.052^*	0.040	0.068
Panel statistic	0.970	−0.827	−1.336	4.437^***	1.674^**	1.034
	(0.166)	(0.796)	(0.909)	(0.000)	(0.047)	(0.151)

Panel stationarity test – first difference of cyclically adjusted expenditure

Countries	Constant			Constant and trend
Countries	k = 1	k = 2	k = 3	k = 1	k = 2	k = 3
Czechia	0.224^**	0.255	0.298	0.140^***	0.142^**	0.137^*
Estonia	0.041	0.105	0.061	0.037	0.043	0.045
Hungary.	0.059	0.141	0.144	0.045	0.141^**	0.131^*
Latvia	0.076	0.075	0.068	0.056^**	0.063	0.068
Lithuania	0.121	0.211	0.194	0.119^***	0.126^*	0.118^*
Poland	0.193^*	0.191	0.198	0.051^*	0.120^*	0.144^**
Slovakia	0.044	0.304	0.204	0.036	0.036	0.063
Slovenia	0.138^*	0.189	0.147	0.108^***	0.137^**	0.074
Bulgaria	0.102	0.073	0.142	0.049^*	0.051	0.061
Romania	0.042	0.147	0.055	0.038	0.048	0.044
Panel statistic	2.237^**	0.668	−0.086	9.258^***	3.130^***	2.196^**
	(0.013)	(0.252)	(0.532)	(0.000)	(0.001)	(0.014)

Panel stationarity test – Revenue

Countries	Constant			Constant and trend
Countries	k = 1	k = 2	k = 3	k = 1	k = 2	k = 3
Czechia	0.146^*	0.314	0.408^*	0.047	0.048	0.068
Estonia	0.157^*	0.271	0.108	0.044	0.066	0.091
Hungary.	0.342^***	0.187	0.148	0.040	0.107^*	0.103
Latvia	0.070	0.448^**	0.313	0.052^*	0.074	0.076
Lithuania	0.074	0.155	0.299	0.051^*	0.076	0.054
Poland	0.170^*	0.409^*	0.352^*	0.065^**	0.101	0.103
Slovakia	0.262^**	0.371^*	0.362^*	0.057^**	0.153^**	0.143^**
Slovenia	0.056	0.128	0.239	0.056^*	0.079	0.074
Bulgaria	0.268^**	0.115	0.121	0.065^**	0.114^*	0.102
Romania	0.074	0.151	0.160	0.049^*	0.121^*	0.116^*
Panel statistic	5.652^***	2.716^***	2.136^***	5.606^***	3.405^***	2.523^***
	(0.000)	(0.003)	(0.016)	(0.000)	(0.000)	(0.006)

Panel cointegration test of revenue and expenditure – Europe

	Z_τ (N)		Z_φ(N)
Models	Value (τ)	P-value	Value (φ)	P-value
No breaks	−1.515	0.065	−0.963	0.168
Level break	0.879	0.810	0.690	0.755
Regime shift	0.422	0.663	0.437	0.669
Number of observations	250		250

Panel cointegration test of cyclically adjusted revenue and cyclically adjusted spending (HP Filter used for detrending series)

	Z_τ(N)		Z_φ(N)
Models	Value (τ)	P-value	Value (φ)	P-value
No breaks	1.731	0.958	−2.409^***	0.008
Level break	−1.984^**	0.024	−2.490^***	0.006
Regime shift	−6.050^***	0.000	−7.182^***	0.000
Number of observations	250		250

Panel stationarity Test test -– First first difference of cyclically adjusted revenue

Countrs	Constant			Constant and trend
Countrs	k = 1	k = 2	k = 3	k = 1	k = 2	k = 3
Czechia	0.060	0.092	0.168	0.058^**	0.068	0.067
Estonia	0.043	0.073	0.097	0.039	0.067	0.086
Hungary.	0.042	0.071	0.121	0.040	0.071	0.122^*
Latvia	0.093	0.096	0.092	0.093^***	0.096	0.088
Lithuania	0.166^*	0.496^**	0.303	0.103^***	0.130^*	0.104
Poland	0.217^**	0.187	0.286	0.033	0.071	0.089
Slovakia	0.102	0.244	0.214	0.102^***	0.104^*	0.095
Slovenia	0.131	0.172	0.185	0.114^***	0.197^**	0.088
Bulgaria	0.1068	0.085	0.117	0.063^**	0.077	0.116^*
Romania	0.120	0.071	0.052	0.039	0.047	0.042
Panel statistic	2.244^**	0.421	0.190	9.430^***	3.306^***	2.275^**
	(0.012)	(0.337)	(0.425)	(0.000)	(0.000)	(0.011)

Summary of existing empirical panel fiscal sustainability test

Reference	Sustainability test	Period and country	Findings
Afonso and Rault (2010)	Stationarity of debt and cointegration between revenue and expenditure	15 Selected EU countries (1970–2006)	Fiscal stance sustainability confirmed
Baldi and Staehr (2015)	Estimated fiscal reaction function of primary balance, debt and business cycle viables	Different groups of EU countries (2001–2004)	Sustainable fiscal stance for all groups post financial crises
Beqiraj et al. (2018)	Panel cointegration test between primary balance and public debt	21 OECD countries (1991–2015)	Fiscal stance judged to be unsustainable
Boekemeier and Stoian (2018)	Fiscal reaction function of primary balance and debt	CEEC (1997–2013)	Fiscal stance sustainable for selected countries
Brady and Magazzino (2018)	Stationarity of public debt	19 European countries (1970–2016)	Fiscal stance sustainability confirmed
Checherita-Wesphal and Žďárek (2017)	Fiscal reaction function of primary balance response to debt	18 Euro Area countries (1970–2013)	Sustainable fiscal stance
Claeys (2007)	Cointegration between revenue, spending and net interest payment	Selected European countries (1970–2001)	Sustainable fiscal policy
Krajewski et al. (2016)	Cointegration between revenue and expenditure and a fiscal reaction function	CEEC (1990–2012)	Sustainable fiscal stance
Lee et al. (2018)	Fiscal reaction function of primary balance response to debt	EU regional groups (1950–2014)	Varied results depending on the region
Llorca and Redzepagic (2008)	Cointegration between revenue and expenditure	CEEC (1999:1–2006:1)	Fiscal stance sustainable in selected countries
Prohl and Schneider (2006)	Cointegration between budget deficit and public debt	15 EU countries (1970–2004)	Fiscal stance sustainability confirmed
Westerlund and Prohl (2010)	Cointegration between revenue and expenditure	8 rich OECD countries (1977:1–2006:4)	Sustainability hypothesis confirmed for selected countries

Breusch–Pagan and Pesaran cross-sectional dependence test

Variables	Breusch–Pagan CD test	Pesaran CD test
Variables	Chi-square, P-value	z-value, P-value
Expenditure to GDP ratio	107.84, 0.000	5.494, 0.000
Revenue to GDP ratio	113.42, 0.000	1.716, 0.0862

Long run coefficient for cyclically adjusted fiscal variables

Stat	MG	CCEMG	MG-FMOLS	MG-DOLS^a
γ	0.499	0.364	0.938	0.935
Test stat	3.390	3.815	480.74	604.07
P-value	0.000	0.000	0.000	0.000
Std error	0.147	0.095	0.002	0.02
Obs	220	220	210	130
Shapiro–Wilk Normality test	0.970	0.977	0.992	0.986
P-value	(0.000)	(0.001)	(0.260)	(0.220)
Peseran CD test	3.652	−2.411	3.328	−0.260
P-value	(0.000)	(0.016)	(0.001)	(0.795)

Panel stationarity test – cyclically adjusted expenditure

Countries	Constant			Constant and trend
Countries	k = 1	k = 2	k = 3	k = 1	k = 2	k = 3
Czechia	0.144^*	0.165	0.132	0.065^**	0.134^**	0.126^**
Estonia	0.080	0.401	0.385^*	0.046	0.105^*	0.092
Hungary.	0.157^*	0.096	0.068	0.059^**	0.100	0.126^**
Latvia	0.106	0.202	0.155	0.045	0.080	0.073
Lithuania	0.089	0.172	0.146	0.072^***	0.135^**	0.126^**
Poland	0.154^*	0.466^**	0.577^**	0.047	0.074	0.073
Slovakia	0.095	0.258	0.206	0.063^**	0.146^**	0.126^**
Slovenia	0.091	0.488^**	0.332	0.070^**	0.140^**	0.093
Bulgaria	0.102	0.072	0.097	0.068^**	0.072	0.065
Romania	0.064	0.087	0.110	0.049^*	0.089	0.110
Panel statistic	2.491^***	2.376^***	1.461^*	7.017^***	4.511^***	2.885^***
	(0.006)	(0.009)	(0.072)	(0.000)	(0.000)	(0.002)

Dates of structural breaks for individual series

Countries (CEEC)	Revenue	Expenditure
Czechia	2002, 2010	2003
Estonia	1998, 2008, 2011	1999, 2007, 2010
Hungary	1997, 2006, 2011, 2015	1998, 2001, 2015
Latvia	1999, 2009	1997, 2000, 2008, 2011
Lithuania	2000	2000, 2008, 2011
Poland	1997, 2008, 2015	1997, 2011
Slovakia	1997, 2000, 2003, 2012	2002, 2008
Slovenia	2011, 2015	2008, 2015
Bulgaria	1997, 2008, 2012	1998, 2002
Romania	1997	1998, 2001, 2006, 2012

Panel Cointegration cointegration test of cyclically adjusted revenue and cyclically adjusted spending

	Z_τ(N)		Z_φ(N)
Models	Value (τ)	P-value	Value (φ)	P-value
No breaks	−2.842^***	0.002	−3.587^***	0.000
Level break	−3.076^***	0.001	−2.950^***	0.002
Regime shift	−2.368^***	0.009	−2.984^***	0.001
Number of observations	220		220

Panel stationarity Test test -– Cyclically cyclically adjusted revenue

Countrs	Constant			Constant and trend
Countrs	k = 1	k = 2	k = 3	k = 1	k = 2	k = 3
Czechia	0.160^*	0.416^**	0.519^**	0.048^*	0.070	0.060
Estonia	0.079	0.414^*	0.249	0.044	0.042	0.089
Hungary	0.040	0.398^*	0.348^*	0.039	0.058	0.103
Latvia	0.037	0.324^*	0.356^*	0.037	0.036	0.051
Lithuania	0.061	0.242	0.165	0.059^**	0.145^**	0.111
Poland	0.241^**	0.434^**	0.435^*	0.078^***	0.056	0.072
Slovakia	0.080	0.262	0.230	0.056^**	0.149^**	0.123^*
Slovenia	0.097	0.237	0.261	0.075^***	0.132^*	0.066
Bulgaria	0.040	0.281	0.343^*	0.037	0.035	0.071
Romania	0.083	0.246	0.338	0.054^*	0.044	0.073
Panel statistic	1.524^***	4.395^***	3.768^***	5.575^***	1.985^**	1.682^**
	(0.064)	(0.000)	(0.000)	(0.000)	(0.023)	(0.046)

Fiscal Sustainability Hypothesis Test in Central and Eastern Europe: A Panel Data Perspective

Figures & Tables

Figure 1

Figure 2

Figure 3

Wald test of coefficient and confidence intervals

Panel summary Statistics

Deficit to GDP ratio

Panel Stationarity stationarity test with sharp breaks - – Robustness robustness check

Panel stationarity test – Expenditure

Panel stationarity Test test -– First first difference of Revenuerevenue

Chow test of poolability of coefficients (5% significance level)

Panel stationarity test – first difference of expenditure

Panel stationarity test – first difference of cyclically adjusted expenditure

Panel stationarity test – Revenue

Panel cointegration test of revenue and expenditure – Europe

Panel cointegration test of cyclically adjusted revenue and cyclically adjusted spending (HP Filter used for detrending series)

Panel stationarity Test test -– First first difference of cyclically adjusted revenue

Summary of existing empirical panel fiscal sustainability test

Breusch–Pagan and Pesaran cross-sectional dependence test

Long run coefficient for cyclically adjusted fiscal variables

Panel stationarity test – cyclically adjusted expenditure

Dates of structural breaks for individual series

Panel Cointegration cointegration test of cyclically adjusted revenue and cyclically adjusted spending

Panel stationarity Test test -– Cyclically cyclically adjusted revenue

Paradigm

My account