Do regional integration and trade linkages promote productivity spillovers? Evidence from the European Union

Figures & data

Table 1. Summary statistics.

Variable	Description	Mean	SD	Min.	Max
y	Real GDP (2005 million US dollars at 2005 PPPs)	518,830.70	813,476.00	5,450.66	4,392,073.00
k	Real capital stock (2005 million US dollars at current PPPs)	2,535,834.00	4,174,380.00	15,276.01	20,900,000.00
l	Average annual hours worked by persons engaged	1,758.69	193.20	1,380.60	2,277.38
exp	Exports of merchandise minus imports of merchandise as a share of GDP i.e. net trade openness	−0.08	0.14	−0.61	0.39
gov	Government spending as a share of GDP	0.21	0.06	0.08	0.46
y	Central and Eastern European Economies (n = 300)	188,365.80	219,988.50	14,015.03	1,211,305.00
k		707,355.70	700,699.90	66,321.37	3,148,161.00
l		1,836.32	121.38	1,592.93	2,115.16
exp		−0.07	0.06	−0.33	0.11
gov		0.27	0.05	0.16	0.46
y	Eurozone Countries (n = 570)	617,395.50	942,330.90	5,450.66	4,392,073.00
k		3,135,475.00	4,829,964.00	15,276.01	20,900,000.00
l		1,804.90	172.67	1,495.46	2,277.38
exp		−0.09	0.16	−0.61	0.39
gov		0.20	0.06	0.08	0.46

N = 810 observations from 27 EU countries from 1990 to 2019. Real GDP, real capital stock and labour employed are reported in levels and are log-transformed and standardized prior to the estimations. Meanwhile, we standardize the time-trend, net exports, and government spending. Standardization (mean 0, standard deviation 1) allows us to interpret the indirect and direct coefficients as elasticities. Finally, the dummy variable cee refers to the Central and European Union Economies and includes Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovenia, and Slovakia. The Eurozone countries are Austria, Belgium, Cyprus, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Portugal, Slovakia, Slovenia, and Spain.

Table 2. Spatial Durbin production frontier coefficients and associated parameters.

Variable	Model Coeff.	Direct Parameter	Indirect Parameter	Total Parameter
K	0.553***	0.557***	0.155	0.711***
	(0.038)	(0.039)	(0.177)	(0.178)
l	0.425***	0.428***	0.529**	0.958***
	(0.037)	(0.037)	(0.232)	(0.226)
k^2	0.132***	0.137***	0.226	0.363*
	(0.026)	(0.025)	(0.206)	(0.203)
l^2	0.020	0.027	0.583**	0.611**
	(0.021)	(0.020)	(0.253)	(0.253)
kl	−0.149***	−0.166***	−1.238***	−1.404***
	(0.043)	(0.041)	(0.449)	(0.448)
t	0.013***	0.013***	0.003*	0.016***
	(0.004)	(0.004)	(0.002)	(0.004)
t^2	0.000***	0.000***	0.000	0.000***
	(0.000)	(0.000)	(0.000)	(0.000)
kt	−0.012***	−0.012***	0.016	0.004
	(0.002)	(0.002)	(0.011)	(0.011)
lt	0.009***	0.009***	−0.003	0.005
	(0.002)	(0.001)	(0.010)	(0.010)
exp	0.039***	0.038***	−0.076**	−0.038
	(0.006)	(0.006)	(0.034)	(0.033)
gov	−0.025***	−0.027***	−0.069***	−0.096**
	(0.006)	(0.005)	(0.025)	(0.024)
cee	−0.161***	−0.153**	0.334	0.180
	(0.061)	(0.061)	(0.439)	(0.403)
Wk	0.014
	(0.142)
Wl	0.347*
	(0.212)
Wk^2	0.157
	(0.165)
Wl^2	0.455**		R^2 within	0.919
	(0.197)
Wkl	−0.962***		R^2 between	0.989
	(0.343)
Wkt	0.015*		R^2 overall	0.986
	(0.008)
Wlt	−0.004		AIC	−1124.65
	(0.008)
Wexp	−0.068**		BIC	−1003.46
	(0.028)
Wgov	−0.051**		Log-likelihood	1.269
	(0.020)
Constant	−0.250*			0.201**
	(0.137)			(0.102)

Definition of variables are in Table 1. *, **, *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. Standard errors are in parenthesis. AIC and BIC denote the Akaike and Bayesian Information Criteria, respectively.

Table 3. Averaged gross efficiency scores across countries.

Country	Total gross efficiency scores	Indirect efficiency scores	Direct efficiency scores
Austria*	110.51%	18.38%	92.13%
Belgium*	112.62%	18.46%	94.16%
Bulgaria^†	97.13%	18.21%	78.91%
Croatia	98.76%	18.44%	80.31%
Cyprus*	104.55%	18.21%	86.34%
Czech*^†	105.33%	18.49%	86.85%
Denmark	110.90%	18.79%	92.12%
Estonia*^†	113.58%	19.04%	94.54%
Finland	113.60%	18.81%	94.79%
France*	113.73%	18.05%	95.68%
Germany*	110.11%	17.81%	92.30%
Greece*	99.66%	18.40%	81.26%
Hungary^†	108.94%	18.42%	90.51%
Ireland*	114.55%	18.80%	95.76%
Italy*	106.35%	18.38%	87.98%
Latvia*^†	116.55%	18.86%	97.69%
Lithuania*^†	112.03%	18.97%	93.06%
Luxembourg*	115.48%	18.87%	96.61%
Malta*	110.27%	18.54%	91.74%
Netherlands*	113.70%	18.42%	95.28%
Poland^†	116.40%	18.44%	97.96%
Portugal*	89.68%	18.53%	71.15%
Romania^†	98.09%	18.44%	79.65%
Slovakia*^†	103.39%	18.46%	84.94%
Slovenia*^†	108.14%	18.36%	89.78%
Spain*	107.36%	18.18%	89.18%
Sweden	114.21%	18.66%	95.55%
Total Average	108.36%	18.50%	89.86%
Total Average Central and Eastern Europe	107.96%	18.57%	89.39%
Total Average Eurozone	108.82%	18.48%	90.34%
Total Average Central and Eastern Europe – Eurozone	109.84%	18.70%	91.14%
Total Average Central and Eastern Europe – Non-Eurozone	105.14%	18.38%	86.76%

Gross time-varying scores are unbounded as it takes into account the own net time-invariant, own net time-varying and gross time-varying efficiency. This is calculated using the spatial multiplier matrix which passes through a provinces’ first order and neighbouring countries and rebounds back to the unit. The scores above 1 suggests that the efficiency spillover is sufficiently large and has pushed the efficiency of the respective unit beyond the best practice frontier. * denotes Eurozone countries and ^† denotes Central and Eastern European countries.

Figure 1. Averaged gross time-varying efficiency across countries. Source: Author’s compilation from Table 3.

Table 4. Averaged TFP growth decomposition across countries.

Country	Total factor productivity	Total technical change	Direct technical change	Indirect technical change	Returns to scale
Austria*	1.084	1.049	1.035	1.036	0.999
Belgium*	1.117	1.051	1.066	1.028	1.038
Bulgaria^†	1.024	1.000	1.024	0.990	1.035
Croatia	1.051	1.033	1.018	1.012	1.006
Cyprus*	1.094	1.096	0.998	1.007	0.992
Czech*^†	1.066	1.021	1.045	1.006	1.039
Denmark	1.067	1.044	1.022	1.022	1.001
Estonia*^†	1.081	1.050	1.031	0.984	1.048
Finland	1.042	1.030	1.012	1.012	1.000
France*	1.078	1.010	1.068	1.031	1.037
Germany*	1.083	0.993	1.090	1.043	1.047
Greece*	1.051	1.035	1.016	1.018	0.998
Hungary^†	1.053	1.019	1.033	0.998	1.035
Ireland*	1.141	1.131	1.010	1.019	0.991
Italy*	1.040	1.012	1.028	1.032	0.995
Latvia*^†	1.092	1.058	1.035	0.989	1.045
Lithuania*^†	1.066	1.030	1.036	0.987	1.049
Luxembourg*	1.121	1.122	0.999	0.999	1.000
Malta*	1.111	1.133	0.978	0.990	0.988
Netherlands*	1.105	1.036	1.070	1.030	1.039
Poland^†	1.050	0.999	1.051	1.004	1.047
Portugal*	1.071	1.044	1.026	1.016	1.010
Romania^†	0.988	0.979	1.009	0.997	1.012
Slovakia*^†	1.062	1.027	1.035	0.995	1.039
Slovenia*^†	1.084	1.048	1.035	0.992	1.043
Spain*	1.071	1.027	1.045	1.027	1.018
Sweden	1.048	1.032	1.016	1.017	0.999
Total Average	1.072	1.041	1.031	1.010	1.020
Total Average Central and Eastern Europe	1.057	1.023	1.033	0.994	1.039
Total Average Eurozone	1.085	1.051	1.034	1.012	1.022
Total Average Central and Eastern Europe – Eurozone	1.075	1.039	1.036	0.992	1.044
Total Average Central and Eastern Europe – Non-Eurozone	1.029	0.999	1.029	0.997	1.032

Total factor productivity growth index is decomposed using a Malmquist index outlined in Section A1. The direct and indirect effect is calculated using the spatial multiplier matrix which passes through a unit’s first order and neighbouring units and rebounds back to the unit. * denotes Eurozone countries and ^† denotes Central and Eastern European countries.

Figure 2. Averaged TFP growth decomposition across countries. Source: Author’s compilation from Table 4.

Figure 3. Averaged TFP growth decomposition over time. Source: Author’s compilation from Table 5.

Table 5. Averaged TFP growth decompositions over time.

Year	Total factor productivity	Total technical change	Direct technical change	Indirect technical change	Returns to scale
1991	1.052	1.022	1.024	1.008	0.999
1992	1.047	1.021	1.023	1.008	0.995
1993	1.086	1.040	1.032	1.009	1.006
1994	1.095	1.039	1.031	1.009	1.017
1995	1.106	1.038	1.030	1.009	1.029
1996	1.096	1.038	1.029	1.009	1.021
1997	1.087	1.037	1.028	1.009	1.014
1998	1.091	1.036	1.027	1.009	1.019
1999	1.084	1.035	1.027	1.009	1.014
2000	1.071	1.034	1.026	1.009	1.002
2001	1.088	1.033	1.025	1.008	1.021
2002	1.069	1.033	1.024	1.008	1.004
2003	1.094	1.032	1.024	1.008	1.030
2004	1.115	1.031	1.023	1.008	1.053
2005	1.089	1.030	1.022	1.009	1.028
2006	1.125	1.030	1.020	1.009	1.065
2007	1.138	1.029	1.019	1.010	1.080
2008	1.139	1.029	1.018	1.011	1.081
2009	1.076	1.028	1.016	1.012	1.020
2010	1.087	1.027	1.015	1.012	1.033
2011	1.091	1.026	1.014	1.013	1.038
2012	1.077	1.026	1.012	1.013	1.026
2013	1.072	1.025	1.011	1.014	1.023
2014	1.076	1.024	1.010	1.014	1.028
2015	1.079	1.023	1.010	1.014	1.032
2016	1.110	1.023	1.009	1.014	1.065
2017	1.072	1.022	1.008	1.014	1.028
2018	1.072	1.021	1.007	1.014	1.030
2019	1.078	1.020	1.007	1.014	1.038
Total Average	1.072	1.041	1.031	1.010	1.020

Total factor productivity growth index is decomposed using a Malmquist index outlined in Section A1. The direct and indirect effect is calculated using the spatial multiplier matrix which passes through a unit’s first order and neighbouring units and rebounds back to the unit. * denotes Eurozone countries and ^† denotes Central and Eastern European countries.