The role of regulation and regional government quality for high-growth firms: the good, the bad and the ugly

Figures & data

Figure 1. European Quality of Government index (QoG), (a) 2010 and (b) 2017.

Note: Darker hues indicate higher quality of government. The legend is based on the quartile calculation for 2010.

Source: Authors’ on calculations based on the European Quality of Government index (QoG) from the QoG Institute, Gothenburg University.

Two maps of Europe displaying sub-national variation in the quality of government in 2010 (left) and 2017 (right).

Figure 2. Share of high-growth firms and quality of regional government, 2017.

Sources: Authors’ on calculations based on the European Quality of Government index (QoG) from the QoG Institute, Gothenburg University; and Eurostat–SBS data.

Scatter plot of regional QoG index and the average regional shares of HGFs in 2017.

Figure 3. Share of high-growth firms by country and sector, average for 2008–18.

Note: The box plots display the first to the third quartiles. The vertical line is the median. Whiskers represent the minimum and maximum.

Countries are: AT = Austria; CZ = Czechia; DK = Denmark; EE = Estonia; ES = Spain; FI = Finland; FR = France; HU = Hungary; IT = Italy; NL = Netherlands; PL = Poland; PT = Portugal; SI = Slovenia; and SK = Slovakia.

Sectors are: B–E: Industry (B = Mining and quarrying; C = Manufacturing; D = Electricity, gas, steam and air-conditioning supply; E = Water supply; sewerage, waste management and remediation activities). F = Construction. G = Wholesale and retail trade; repair of motor vehicles and motorcycles. H = Transport and storage. I = Accommodation and food service activities. J = Information and communication. K–L: K = Financial and insurance activities; L = Real estate activities. M–N: M = Professional, scientific and technical activities; N = Administrative and support service activities. P–Q: P = Education; Q = Human health and social work activities. R–S: R = Arts, entertainment and recreation; and S = Other service activities.

Source: Authors’ calculations based on Eurostat–SBS data.

Two box plots reporting the distribution of HGFs shares by country (within-country variation) and by sector (within-sector variation) of our sample.

Table 1. Description of the variables.

Variable	Description	Sources	Sample
HGF (%)	Number of high-growth enterprises measured in employment (growth by 20% or more for at least three consecutive years) divided by the population of active enterprises	Eurostat-SBS (BD_HGNACE2_R3)	$\mathrm{rjt}$ (2008–18)
QoG	The European Quality of Government index is based on a large citizen survey where respondents are asked about perceptions and experiences with public sector corruption, along with the extent to which citizens believe various public sector services are impartially allocated and of good quality. The index ranges from 0 to 100 (highest quality). It is available only for the survey years 2010, 2013, 2017 and 2021. We use either interpolated data for the missing years, or assume the index to be constant between survey years	The Quality of Government (QoG) Institute, Gothenburg University	$\mathrm{rt}$ (2008–17)
PMR	Impact of the regulatory set-up in energy, e-communications and transport sectors ($k$) on 37 industries ($j$) that use the output of these sectors as intermediate inputs ($w_{\mathrm{jk}}$):	OECD REGIMPACT	$\mathrm{cjt}$ (2008–18)
	$\mathrm{REGIMPAC}{T}_{\mathrm{jt}}={\sum }_{k=1}^K\mathrm{PM}{R}_{\mathrm{kt}}{w}_{\mathrm{jk}}$
L flex	Hiring and firing practice index based on the calculated score from the answers of a representative sample of business leaders in their respective countries to the following question: ‘In your country, to what extent do regulations allow flexible hiring and firing of workers? [1 = not at all; 7 = to a great extent]’. The index has been rescaled to 0–100	The Global Competitiveness Index Historical Dataset – World Economic Forum (WEF)	$\mathrm{ct}$ (2007–17)
Access2K	We built a principal component-based weighted index from three separate scores^a to the survey questions related to (1) ease of access to bank loans, (2) access to equity funding for financing innovative and risky projects and (3) access to finance by issuing bonds or shares on the capital market. The index has been rescaled to 0–100	The Global Competitiveness Index Historical Dataset – WEF	$\mathrm{ct}$ (2008–17)
Churn	The sum of the entry and exit rates, measured as the share of the number of firms’ births and deaths over the total population of active enterprises	Eurostat–SBS (BD_HGNACE2_R3)	$\mathrm{rjt}$ (2008–18)
HT empl	Employment in technology and knowledge-intensive sectors as a percentage of total employment	Eurostat (HTEC_EMP_REG2)	$\mathrm{rt}$ (2008–18)
$\mathrm{\Delta }$GDP	Growth rate of gross domestic product (GDP) per inhabitant at purchasing power standard	Eurostat (NAMA_10R_2GDP)	$\mathrm{rt}$ (2008–18)
literacy1930	Share of the population able to read and write in 1930. Data are available only for 100 European regions	Diebolt and Hippe’s (2019) based on based on Kirk’s (1969) data	$r$ (1930)
HC	Number of scientists and engineers aged 15–74 years as a share of the active population	Eurostat–Human Resources in Science and Technology (HRST)	$\mathrm{rt}$ (2008–18)

Note: $r$, Region; $c$, country; $j$, sector; $t$, year.

^aOur preliminary analysis uses the three indicators separately. However, they all yield similar results, which are available from the authors upon request.

Table 2. Summary statistics.

Variable	Mean	Median	IQR	Minimum	Maximum	N
HGF (%)	0.53	0.37	0.54	0	3.3	9254
QoG	58	57	31	1.8	100	13,820
PMR	4.3	1.8	1.6	0.13	48	14,500
L flex	27	26	19	0	100	14,500
Access2K	42	42	33	0	100	14,500
HC	5	4.7	2.6	1.6	14	11,760
Churn	18	17	7.4	5.1	52	8935
HT empl	3.3	2.9	1.9	0.8	11	12,870
$\mathrm{\Delta }$GDP	1.5	1.9	3.6	–14	12	14,300
literacy1930	79	88	32	29	100	7800

Note: Averages across years, regions (or countries) and sectors.

IQR, interquartile range.

Table 3. Cross-correlation table.

Variables	HGF	QoG	PMR	L flex	Access2K	HC	Churn	HT empl	$\mathrm{\Delta }$gdp
HGF (%)	1.000
QoG	0.192	1.000
PMR	0.332	0.079	1.000
L flex	0.197	0.524	0.051	1.000
Access2K	0.293	0.711	0.082	0.452	1.000
HC	0.168	0.325	0.055	0.432	0.400	1.000
Churn	–0.117	–0.033	–0.203	0.004	0.088	0.080	1.000
HT empl	0.184	0.264	0.039	0.252	0.340	0.750	0.018	1.000
$\mathrm{\Delta }$gdp	0.119	–0.020	0.002	0.299	0.094	0.087	–0.064	0.044	1.000
literacy1930	0.132	0.423	0.012	0.314	0.278	0.292	–0.412	0.466	0.125

Table 4. Averages of high-growth firms (HGFs) (%), product and labour market regulation, by regions group.

QoG	Variable	$\mathrm{\Delta }\mathrm{Qo}{G}_{10-17}<0$	$\mathrm{\Delta }\mathrm{Qo}{G}_{10-17}\ge 0$	Row average	N
High	HGF (%)	0.71	0.27^a	0.67	2315
	PMR	3.76	4.62	3.83
	L flex	38.67	27.90	37.77
Medium	HGF (%)	0.51	0.45	0.50	4421
	PMR	4.63	4.89	4.65
	L flex	21.08	32.04	21.86
Low	HGF (%)	0.42	0.59	0.43	2438
	PMR	4.08	4.87	4.21
	L flex	23.68	33.02	25.28
Column average	HGF (%)	0.53	0.42	0.53
	PMR	4.23	4.82	4.33
	L flex	26.24	31.57	26.77
	N	8721	453	9174

Note: The three levels of the European Quality of Government index (QoG) are defined using the regional average 25th and 75th percentiles. Variables’ averages are across all sectors, years and within groups of regions. The number of observations refers to the minimum available for estimations.

^aOnly three regions in the category high quality of government experienced an increase in quality over the period 2010–17: the small Finnish island Åland and the Spanish Autonomous cities of Ceuta and Melilla. These averages cannot be compared with those of other groups of regions.

Table 5. Instrumental variables (IV) estimation (two-step generalized method of moments – GMM).

Dep. var.: HGFs (%)	(GMM)	(GMMNI)	(GMM)	(GMM)
QoG	0.003***	0.003***	0.005***	0.002***
	(0.000)	(0.000)	(0.000)	(0.000)
PMR	–0.007***	–0.007***	–0.011***	–0.012***
	(0.003)	(0.003)	(0.003)	(0.004)
L flex	0.001**	0.001**	0.009***	0.001**
	(0.000)	(0.000)	(0.001)	(0.000)
HC	0.017***	0.016***	0.028***	0.017***
	(0.004)	(0.004)	(0.004)	(0.004)
Access2K	0.001***	0.001***	0.001***	0.001***
	(0.000)	(0.000)	(0.000)	(0.000)
Churn	0.003***	0.003***	0.002**	0.004***
	(0.001)	(0.001)	(0.001)	(0.001)
HT empl	0.007*	0.007*	–0.000	0.007*
	(0.004)	(0.004)	(0.004)	(0.004)
$\mathrm{\Delta }$GDP	0.013***	0.013***	0.009***	0.013***
	(0.002)	(0.002)	(0.002)	(0.002)
L flex#QoG			–0.001***
			(0.000)
PMR#QoG				0.001*
				(0.000)
R^2	0.572	0.571	0.578	0.572
N	5578	5578	5578	5578
Sargan	163.181***	167.984***	138.565***	163.889***
Weak identification F	10,839.445	6041.122	1146.245	6059.654
Endogeneity test	11.143**	13.962***	13.587***	10.834**

Note: Significance: p < 0.01***, p < 0.05**, p < 0.1*. Robust standard errors are shown in parentheses. All right-hand-side variables are lagged by one year. All econometric specifications include year, sector and region dummies.

N, number of observations.

Table 6. Literacy rate as an instrumental variable (IV).

Dep. var.:	QoG	HGFs (%)	HGFs (%)
literacy1930	0.101***
	(0.010)
QoG		–0.001
		(0.001)
$Q\hat{o}G$			0.002**
			(0.001)
PMR	–0.018	–0.177***	–0.187***
	(0.034)	(0.053)	(0.039)
L flex	0.091***	0.003*	0.004**
	(0.010)	(0.002)	(0.002)
HC	0.295***	0.013*	0.015**
	(0.081)	(0.007)	(0.006)
Access2K	0.076***	0.007***	0.006***
	(0.009)	(0.000)	(0.000)
Churn	–0.018	0.028***	0.025***
	(0.015)	(0.002)	(0.002)
HT empl	0.177	0.016*	0.022**
	(0.143)	(0.009)	(0.010)
$\mathrm{\Delta }$GDP	–0.241***	–0.000	0.001
	(0.029)	(0.002)	(0.002)
R^2	0.969	0.190	0.161
N	5618	4411	3848

Note: The sample used for these estimations includes only regions available in both the Eurostat and the literacy rate datasets (Diebolt & Hippe, 2019), which are regions of Austria, Spain, Finland, France, Italy and Portugal. The first column reports the results obtained from a first regression where we isolate the source of variation of institutional quality explained by literacy rate in 1930. Although not reported, the first regression includes region, sector and year dummies.

Significance: p < 0.01***, p < 0.05**, p < 0.1*. Robust standard errors are shown in parentheses. All right-hand-side variables are lagged by one year. All econometric specifications include year, sector and region dummies.

N, number of observations.

Table 7. Instrumental variables (IV) generalized method of moments (GMM) estimation, by level of the European Quality of Government index (QoG).

Dep. var.: HGFs (%)	High QoG	Medium QoG	Low QoG
PMR	–0.015	–0.111*	–0.050***
	(0.054)	(0.065)	(0.015)
L flex	0.003	0.006***	0.001
	(0.005)	(0.002)	(0.001)
HC	0.067***	0.000	–0.012
	(0.008)	(0.010)	(0.012)
Access2K	0.003***	0.007***	0.002***
	(0.001)	(0.001)	(0.000)
Churn	0.030***	0.029***	0.003**
	(0.004)	(0.003)	(0.001)
HT empl	–0.020	0.054***	–0.025
	(0.014)	(0.019)	(0.015)
$\mathrm{\Delta }$GDP	–0.007***	0.002	0.009***
	(0.003)	(0.003)	(0.002)
RMSE	0.209	0.192	0.148
R^2	0.142	0.306	0.075
N	1396	2779	1480
Cragg–Donald Wald F	26.904	8.640	8.828

Note: Significance: p < 0.01***, p < 0.05**, p < 0.1*. Robust standard errors are shown in parentheses. All right-hand-side variables are lagged by one year. All econometric specifications include year, sector and region dummies.

N, number of observations; RMSE, root mean square error.

Figure 4. European Quality of Government index (QoG), by country, 2010 (vertical axis) and 2017 (horizontal axis).

Source: Authors’ calculations based on the European Quality of Government index (QoG) from the QoG Institute, Gothenburg University.

For each country, a scatter plot of regional QoG indices in 2010 (vertical axis) and in 2017 (horizontal axis), and a 45 degree line across each plot.

Table 8. Instrumental variables (IV) generalized method of moments (GMM) estimation, by European Quality of Government index (QoG) dynamics.

Dep. var.: HGFs (%)	$\mathrm{\Delta }\mathrm{Qo}{G}_{2010-2017}<0$	$\mathrm{\Delta }\mathrm{Qo}{G}_{2010-2017}\ge 0$
PMR	–0.048***	–0.044
	(0.016)	(0.077)
L flex	0.004***	0.005
	(0.001)	(0.004)
HC	0.035***	0.084***
	(0.004)	(0.029)
Access2K	0.004***	–0.001
	(0.000)	(0.002)
Churn	0.017***	0.011**
	(0.001)	(0.005)
HT empl	–0.001	0.063***
	(0.008)	(0.022)
$\mathrm{\Delta }$GDP	0.002**	0.012**
	(0.001)	(0.006)
R^2	0.127	0.493
N	5535	120

Note: Significance: p < 0.01***, p < 0.05**, p < 0.1*. Robust standard errors are shown in parentheses. All right-hand-side variables are lagged by one year. All econometric specifications include year, sector and region dummies.

N, number of observations.

Table 9. Fixed effects estimation results.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
QoG	–0.004***	–0.003***	–0.003***	–0.000	–0.001**	–0.001*	–0.001	–0.001
	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
PMR		–0.029***	–0.029***	–0.022***	–0.024***	–0.031***	–0.024***	–0.024***
		(0.006)	(0.006)	(0.006)	(0.006)	(0.007)	(0.006)	(0.007)
L flex			0.001**	0.006***	0.006***	0.005***	0.005***	0.005***
			(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
HC				0.025***	0.028***	0.036***	0.037***	0.037***
				(0.006)	(0.006)	(0.006)	(0.006)	(0.006)
Access2K					0.002***	0.003***	0.002***	0.002***
					(0.000)	(0.000)	(0.000)	(0.000)
Churn						0.014***	0.014***	0.014***
						(0.001)	(0.002)	(0.002)
HT empl							–0.004	–0.005
							(0.009)	(0.010)
$\mathrm{\Delta }$GDP								–0.000
								(0.001)
R^2	0.024	0.030	0.031	0.053	0.070	0.097	0.093	0.093
N	9302	9302	9302	7916	7916	7111	6688	6648

Note: Significance: p < 0.01***, p < 0.05**, p < 0.1*. Robust standard errors are shown in parentheses. All right-hand-side variables are lagged by one year. All econometric specifications include a constant, not reported.

N, number of observations.

Diebolt, C., & Hippe, R. (2019). The long-run impact of human capital on innovation and economic development in the regions of Europe. Applied Economics, 51(5), 542–563. https://doi.org/10.1080/00036846.2018.1495820

 Web of Science ®Google Scholar