Digital Twin Readiness and Green Entrepreneurship: Evidence from the Baltic States

Variable	Measurement	Unit	Level	Source
Startups_ict	Newly established cleantech and environmentally oriented start-ups	Count	Industry–country–year	Dealroom.co; PitchBook; regional cleantech reports
VC_ict	Venture-capital and private-equity funding raised by green start-ups	Constant 2015 EUR	Industry–country–year	Dealroom.co; PitchBook; regional cleantech reports; Eurostat HICP
Patents_ict	Patent applications in CPC Y02 and related climate-technology subclasses	Count	Industry–country–year	OECD Patent Statistics Database; EPO PATSTAT
DTRI_ict	DTRI (first principal component of AI, IoT, cloud, big data and ICT specialists)	PCA index	Industry–country–year	Eurostat ICT survey
GVA_ict	Gross value added	EUR	Industry–country–year	Eurostat National Accounts
GFCF_ict	Gross fixed capital formation	EUR	Industry–country–year	Eurostat National Accounts
EMP_ict	Employment	Tens of thousands of persons	Industry–country–year	Eurostat National Accounts
GDP pc_ict	GDP per capita (PPP)	Constant international dollars	Country–year	World Bank; IMF
RnD_ict	R&D expenditure	% of GDP	Country–year	Eurostat; OECD
EDU_ict	Population aged 25–34 with tertiary education	%	Country–year	Eurostat

[i] Source: Authors' elaboration based on Dealroom.co, PitchBook, regional cleantech reports, OECD Patent Statistics Database, EPO PATSTAT, Eurostat, WIPO, World Bank, IMF and OECD.

[ii] Note: Dependent variables are entered as ln(1 + Y) in the regressions.

[iii] AI, artificial intelligence; DTRI, digital twin readiness index; ICT, information and communication technology; EUR, euro; GDP, gross domestic product; HICP, Harmonised Index of Consumer Prices; IMF, International Monetary Fund; IoT, Internet of Things; OECD, Organisation for Economic Co-operation and Development; PATSTAT, Worldwide Patent Statistical Database of the European Patent Office; PCA, principal component analysis; PPP, purchasing power parity; R&D, research and development.

Table 2.

PCA construction of the DTRI

Indicator	Mean	SD	PCA coefficient	Factor loading	Direction
AI use	8.16	7.891	0.518	0.912	Positive
IoT use	26.075	10.132	0.056	0.098	Positive
Cloud computing	28.575	16.036	0.403	0.71	Positive
Big data analytics	11.651	5.21	0.519	0.914	Positive
ICT specialists	22.374	20.008	0.544	0.959	Positive
Explained variance of PC1				62.03%

[i] Note: All indicators are standardised over the pooled country–industry–year sample before PCA. PCA coefficients refer to the eigenvector weights used to construct the first principal component. Factor loadings report the correlation between each standardised indicator and the DTRI score. The sign of PC1 is normalised so that higher values indicate stronger digital twin-related readiness.

[ii] AI, artificial intelligence; DTRI, digital twin readiness index; ICT, information and communication technology; SD, standard deviation; IoT, Internet of Things; PC1, first principal component; PCA, principal component analysis.

Table 3.

Descriptive statistics for key variables, 2015–2024

Variables	Observations (N)	Mean	SD	Minimum	Maximum
*Y_ict*
Startups_ict (number)	270	0.61	1.23	0	5
VC_ict (millions of euros)	270	45.89	203.15	0	2679.5
Patents_ict (number)	270	5.38	9.01	0	42
X_ict
DTRI_ict (index)	270	−0.16	1.45	−2.38	6.08
Industry-level control variable
GVA_ict (billions of euros)	270	12.37	15.81	0.03	70.92
GFCF_ict (billions of euros)	270	2.95	4.12	0.03	16.88
EMP_ict (tens of thousands of people)	270	2.51	2.68	0.03	10.94
Country-level control variable
GDP_pc_ict (thousands of constant international dollars)	270	18.35	4.87	12.01	28.74
RnD_ict (% of GDP)	270	0.46	0.29	0.11	1.06
EDU_ict (% of pop. 25–34)	270	47.11	6.78	38.5	58.2

[i] Source: own calculations based on the dataset described in Table 1; Eurostat ICT usage and e-commerce survey indicators (for DTRI components) and Eurostat National Accounts (industry controls), with country-level indicators from World Bank/IMF and Eurostat/OECD.

[ii] Note: Boldface identifies DTRI, the focal explanatory variable.

[iii] DTRI, digital twin readiness index; ICT, information and communication technology; SD, standard deviation; EMP, employment; GFCF, gross fixed capital formation; GDP, gross domestic product; GVA, gross value added; IMF, International Monetary Fund; OECD, Organisation for Economic Co-operation and Development; R&D, research and development; VC, venture capital.

Trends in Baltic green entrepreneurial outcomes, 2015–2024. *Source:* Authors' calculations based on the country–industry–year panel described in Section 3 and Table 1

Average DTRI in the Baltic States, 2015–2024. *Source:* Own calculations based on Eurostat ‘Enterprise use of ICTs and e-commerce’ survey indicators. DTRI, digital twin readiness index; ICT, information and communication technology

Table 4.

Baseline fixed-effects regression results for green entrepreneurial outcomes

Variables	(1) ln(1 + Startups_ict)	(2) ln(1 + VC_ict)	(3) ln(1 + Patents_ict)
DTRI	0.079^*	0.142^**	0.103^**
	(0.042)	(0.061)	(0.049)
ln(GVA)	0.195^**	0.321^***	0.174^*
	(0.088)	(0.105)	(0.096)
ln(GFCF)	0.091	0.182^*	0.088
	(0.071)	(0.095)	(0.080)
GDP_pc_ct	0.488^**	0.815^***	0.601^**
	(0.225)	(0.287)	(0.271)
R&D	1.192^**	2.433^***	1.895^***
	(0.510)	(0.721)	(0.654)
Fixed effects
Industry	Yes	Yes	Yes
Country	Yes	Yes	Yes
Year	Yes	Yes	Yes
Statistic
Observations (N)	270	270	270
Within-group R-squared	0.27	0.34	0.30

Source: Own calculations based on the dataset described in Table 1 (2015–2024).

Note: Boldface identifies DTRI, the focal explanatory variable. Robust standard errors clustered at the country–industry level are reported in parentheses.

* p < 0.10,

** p < 0.05,

*** p < 0.01.

DTRI, digital twin readiness index; GFCF, gross fixed capital formation; GDP, gross domestic product; GVA, gross value added; R&D, research and development; VC, venture capital.

Table 5.

Difference-in-differences estimates around digitalisation policy exposure

Dependent Variable	ATT
ln(1 + Startups_ict)	0.168^*
	(0.091)
ln(1 + VC_ict)	0.275^**
	(0.113)
ln(1 + Patents_ict)	0.221^**
	(0.098)

Source: Own calculations based on the dataset described in Table 1.

Note: Robust standard errors in parentheses.

* p < 0.1,

** p < 0.05.

ATT, average treatment effect on the treated; VC, venture capital.

Event-study estimates around digitalisation policy exposure and green entrepreneurial outcomes Source: Own calculations based on the dataset described in Table 1

Table 6.

Instrumental-variable regression results

	(1) ln(1 + Startups_ict)	(2) ln(1 + VC_ict)	(3) ln(1 + Patents_ict)
Panel A: First-stage regression (dependent variable: DTRI)
Instrumental variable (IV)	0.812^***	0.812^***	0.812^***
	(0.185)	(0.185)	(0.185)
Kleibergen–Paap F-statistic	19.24	19.24	19.24
Panel B: Second-stage regression (dependent variable: green entrepreneurial outcomes)
DTRI (instrumental variables approach)	0.181^*	0.315^***	0.266^***
	(0.098)	(0.109)	(0.092)
Statistic
Observations (N)	270	270	270

Source: Own calculations based on the dataset described in Table 1.

Note: Robust standard errors clustered at the country–industry level are reported in parentheses.

* p < 0.10,

*** p < 0.01.

DTRI, digital twin readiness index; VC, venture capital.

First-stage relationship in the instrumental-variable specification.
Source: Authors' calculations based on the country–industry–year panel described in Section 3 and Table 1

Comparison of DTRI coefficients across empirical models
Source: Authors' calculations based on the regression results reported in Tables 4–6. DTRI, digital twin readiness index

Table A.1.

FE and IV regression results for the later-period subsample, 2021–2024

Variables	(1) ln(1 + Startups) (FE)	(2) ln(1 + VC) (FE)	(3) ln(1 + Patents) (FE)	(4) ln(1 + Startups) (IV)	(5) ln(1 + VC) (IV)	(6) ln(1 + Patents) (IV)
DTRI	0.091^*	0.137^**	0.115^*	0.195^*	0.302^**	0.255^**
	−0.052	−0.065	−0.061	−0.105	−0.122	−0.114
Controls	Yes	Yes	Yes	Yes	Yes	Yes
Fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Observations	108	108	108	108	108	108
K-P F-stat				15.88	15.88	15.88

Note: Robust standard errors clustered at the country–industry level are reported in parentheses. The subsample contains 108 country–industry–year observations from 2021 to 2024. Reported controls follow the baseline specification: ln(GVA), ln(GFCF), GDP per capita and R&D intensity. The Kleibergen–Paap F-statistic is reported only for the IV specifications.

* p < 0.10,

** p < 0.05.

DTRI, digital twin readiness index; FE, fixed effects; GFCF, gross fixed capital formation; GDP, gross domestic product; GVA, gross value added; IV, instrumental variable; R&D, research and development; VC, venture capital.

References

Authors

Metrics

Articles in this issue

DOI: https://doi.org/10.2478/ceej-2026-0017 | Journal eISSN: 2543-6821 | Journal ISSN: 2544-9001

Journal RSS Feed

Language: English

Page range: 304 - 326

Submitted on: Jan 10, 2026

Accepted on: Jul 10, 2026

Published on: Aug 7, 2026

Published by: Faculty of Economic Sciences, University of Warsaw

In partnership with: Paradigm Publishing Services

Keywords:

Central and Eastern Europe,

digital readiness,

digital twins,

green entrepreneurship,

post-transition economies

JEL:

O33,

Q55,

L26

Related subjects:

Business and economics,

Political economics,

Economic theory, systems and structures,

Microeconomics,

Macroeconomics,

Economic policy

© 2026 Weibo Zhou, Anatolijs Krivins, published by Faculty of Economic Sciences, University of Warsaw
This work is licensed under the Creative Commons Attribution 4.0 License.

Volume 13 (2026): Issue 60 (January 2026)

Digital Twin Readiness and Green Entrepreneurship: Evidence from the Baltic States

Figures & Tables

Table 1.

Table 2.

Table 3.

Figure 1.

Figure 2.

Table 4.

Table 5.

Figure 3.

Table 6.

Figure 4.

Figure 5.

Table A.1.

Paradigm

My account