Examining the Nexus Between Fiscal Decentralization, Green Finance, and the Digital Economy: A Cross-Country Panel Study

Abstract

This paper explores the relationship between fiscal decentralization, green finance, and the digital economy in driving sustainable development, using a balanced cross-country panel dataset spanning 2014–2022, for 29 European countries. Employing dynamic panel estimation techniques, including system generalized method of moments (GMM), the research investigates how fiscal decentralization, green finance, and the digital economy (each of them individually and through interaction mechanisms), dynamically shape sustainable development performance in the presence of endogeneity and temporal persistence. The findings reveal strong inertia in sustainable development, which depends on its previous level. Fiscal decentralization has complex effects: revenue autonomy supports sustainability, whereas expenditure autonomy may undermine it, suggesting differences in how resources are used efficiently at the local versus central levels. Digitalization acts as a catalyst, boosting the effectiveness of environmental taxes and enhancing local spending outcomes. However, if fiscal administrations are not digitally integrated, digitalization may weaken the benefits of decentralized revenues. This study advances the literature by integrating fiscal, financial, and digital views, providing new insights into policy coordination.

1. Introduction

It is the common goal of all the European Union member states to promote sustainable development and ensure a cleaner, greener environment (Satrovic et al., 2024) through the political guidelines covering priorities (European Commission, 2019) such as the European Green Deal, a Europe fit for the digital age, and an economy that works for people. European Union member states have adopted a democratic model that reallocates administrative and financial responsibilities from central authorities to local governments. This process aims to give local governments increased authority over their fiscal policies, budgets, and resource allocations (Satrovic et al., 2024). This shift provides the foundation for a decentralization approach at the local level, emphasizing integrated responsibilities across national and international policies, particularly in environmental protection and digitalization. Sustainable finance involves supporting both current environmentally friendly practices (green finance) and those transitioning toward greener standards over time (transition finance) (European Commission, 2025). While the national-level fiscal decentralization framework resembles this model, outcomes vary depending on the level of fiscal discipline and administrative capacity.

There is a scarcity of studies examining the potential of fiscal decentralization, green finance, and the digital economy as a “triple nexus” that collectively shape the pace and direction of growth and sustainable development. This study provides new perspectives and insights into the potential interconnections among fiscal decentralization, green finance, and the digital economy in shaping growth and sustainable development across 29 European countries, using a balanced cross-country panel dataset spanning 2014–2022. The key question is whether each factor—fiscal decentralization, green finance, and the digital economy—or their interactions influence the level of sustainable development. The novelty of the study lies in the development of a composite sustainability index used as the dependent variable in relation to fiscal decentralization, green finance, and the digital economy. Another innovation is the development of independent interaction variables that better illustrate the combined impact on sustainability, making them more informative for economic policy.

The paper is organized into six sections. Section 2 presents the theoretical approach to the relationship among fiscal decentralization, green finance, and the digital economy in promoting sustainable development. Section 3 explains the methodology. Section 4 presents the GMM-based models. Section 5 discusses the findings, and Section 6 summarizes the conclusions.

2. Theoretical Framework

The role of fiscal decentralization is to enhance local governance, improve public service delivery, and foster economic growth (Khan et al., 2021). Fiscal decentralization reshapes local government incentives and budgetary authority, mobilizing capital for environmentally friendly projects in ways that support green finance. The integration of environmental and climate change considerations into EU policies also creates greening opportunities for budget support and local-level investments. The digital economy facilitates information flows, reduces transaction costs, and supports innovative financing mechanisms. Together, these dimensions constitute a “triple nexus” that affects how countries pursue low-carbon and inclusive growth (Z. Liu et al., 2024; Nepal et al., 2024; Zhan et al., 2022) toward the main aim of society’s sustainable development.

The impact of fiscal decentralization on environmental quality has mixed results in the literature. R. Liu et al. (2022) identified a U-shaped nonlinear relationship between, on the one hand, the green development efficiency, and on the other hand, revenue and expenditure decentralization, as calculated by the superefficiency Slack-Based Measure (SBM) model. First, the effect of fiscal decentralization on green development efficiency is negative but is mitigated as local governments’ environmental preferences increase. This suggests that the effectiveness of decentralization depends on both institutional maturity and environmental accountability mechanisms. Further, Zahra and Badeeb (2022), using the Nonlinear Autoregressive Distributed Lag (NARDL) approach across five OECD countries, found that the ecological footprint responds asymmetrically to positive and negative fiscal decentralization in both the short and long run. In countries with strong institutional capacity, decentralization fosters sustainability, but in others with low institutional capacity, it exacerbates ecological degradation.

Zhan et al. (2022) used a combined methodology based on the Slack-Based Measure (SBM) model, often the Super-SBM version, for static efficiency measurement and the Global Malmquist–Luenberger (GML) index for dynamic productivity analysis, particularly for green total factor productivity (SBM-GML model) to evaluate green total factor productivity across 30 provinces in China, finding that financial decentralisation has significantly weakened the increase in green total factor productivity because the local government’s financial and administrative power is fully matched, being one of the main reasons financial decentralisation shows the inhibitory effect on green total factor productivity. The authors identified some policy recommendations that actively promote the positive regulatory effect of local government competition to enhance the green total factor productivity.

F. Wang et al. (2022), using Methods of Moments Quantiles Regression (MMQR) for seventeen developed countries, confirm the asymmetric effect of fiscal decentralization, green technology innovation, and institutional efficiency on carbon emissions, and validate that their effect is not alike across all distributions, rather significantly varied at lower, medium, and higher quantiles.

Overall, these findings indicate that the environmental effects of fiscal decentralization are context-dependent, shaped by sub-national governance quality and maturity, regulatory frameworks, and national and sub-national political incentives.

Green finance has become a crucial instrument for channeling capital toward achieving the Sustainable Development Goals (SDGs) and fostering resilient growth. Countries with stronger green financing performance have fewer environmental issues (Iqbal et al., 2021). Green finance plays a significant role in green growth and transition to a green economy (Bhatnagar et al., 2024; Cao et al., 2021; Desalegn & Tangl, 2022; Fankhauser et al., 2013; Hung, 2023; Tolliver et al., 2021).

Nepal et al. (2024) find a favorable impact of international green financing on low-carbon energy transformation, with this effect particularly evident for hydro and wind energy consumption in developing countries, using multiple econometric models, such as Ordinary Least Squares and a Fixed-Effects model. Behera et al. (2024) examined ten selected European Union nations from 2000 to 2020 based on multiple estimations using the cross-sectional dependence and slope coefficient homogeneity by Pesaran & Yamagata, cross-sectional augmented ImPesaran (CIPS), and cross-sectional augmented Dickey–Fuller (CADF), reporting that both green finance and fiscal decentralization significantly influence renewable energy use, whereas political risk impedes it.

Zhang et al. (2025) investigated the impact of green municipal bond issuance on air quality in China, using a staggered difference-in-differences (DID) approach with data from 2014 to 2023, and found that such issuance significantly improves local air quality. Further analysis indicated that this effect operates through investments in infrastructure development and signalling effects.

The rise of the digital economy in the knowledge society, encompassing digital finance, big data, artificial intelligence, and internet-based platforms, has transformed the landscape of green development. Digitalization and digital transformation affect the effectiveness of governments in managing public funds to meet the needs of citizens and economic agents (Androniceanu & Georgescu, 2023; Androniceanu et al., 2022). Z. Liu et al. (2024), employing two-way Fixed-Effects (FE) models, instrumental variables, and spatial econometric techniques on data from 30 provinces and cities in China from 2004 to 2019, found that the advancement of the digital economy significantly enhances the quality and efficiency of green innovation. Their study revealed a spatial demonstration effect, showing that fiscal decentralization enhances the impact of the digital economy in neighboring regions.

Recent research has begun to analyze the interconnectedness of these three domains. Gariba et al. (2024) based their study on a panel dataset from 2016 to 2022, used a quantitative research design, and applied structural equation modeling (PLS-SEM), finding that fiscal decentralization has a significant negative effect on economic sustainability but a significant positive effect on environmental and social SDGs. In addition, fiscal decentralization has a significant positive effect on the digital economy. Their study also found that the digital economy has a significant positive relationship with economic and social SDGs, and a negative but significant influence on environmental sustainability. This study also demonstrated that the digital economy mediates the relationship between fiscal decentralization and sustainability.

C. Wang et al. (2024) used a bidirectional fixed-effects model, baseline regression, and robustness tests to analyze data from 31 provinces in China between 2012 and 2021. They found that the development of China’s digital economy improves the level of the green economy. Another result is the influence of the digital economy on the green economy, in which industrial structure upgrading and technological innovation play significant roles. Nepal et al. (2024) found that the digital economy significantly contributes to driving the low-carbon energy transition and enhances the role of green finance therein. X. Liu (2025), based on provincial panel data from China, employs spatial econometric models and Data Envelopment Analysis (DEA) methods to systematically investigate the impact of digital financial development on Green Total Factor Productivity (GTFP) and its spatial correlation characteristics. It was found that digital finance not only significantly enhances local green production efficiency, but also creates positive spatial spillover effects on neighboring regions through technology diffusion and factor mobility.

Despite substantial progress, significant research is still needed. The triangular nexus, which integrates all three dimensions into a single empirical model, remains rare because most studies examine pairwise relationships. However, the literature demonstrates that fiscal decentralization, green finance, and the digital economy are interlinked in shaping pathways toward sustainable development. Fiscal decentralization provides local budgetary capacity and incentives; green finance channels resources to environmentally sustainable sectors; and the digital economy amplifies both through innovation, transparency, and efficiency. Yet, the success of this nexus depends on the institutional quality that governs local fiscal behavior, financial regulation, environmental protection, and digital inclusion. Future empirical studies that integrate these three elements can provide crucial insights for designing coordinated policies to achieve the Sustainable Development Goals and a low-carbon digital future.

To fill gaps in existing research, this paper offers new perspectives and insights into the potential links among fiscal decentralization, green finance, and the digital economy in shaping growth and sustainable development across 29 European countries, based on the main question is whether each factor—fiscal decentralization, green finance, and the digital economy—or their interactions affect the level of sustainable development.

The main hypotheses of this study are as follows:

H1: 

Sustainable development is influenced by fiscal autonomy, digitalization and green policies.

H2: 

Models with larger interactions and lags are more informative for economic policy.

The study’s novelty lies in developing a composite sustainability index, used as the dependent variable, related to fiscal decentralization, green finance, and the digital economy. Additionally, it introduces independent interaction variables that demonstrate the combined impact on sustainability.

3. Materials and Methods

This study investigates the interconnected roles of fiscal decentralization, green finance, and the digital economy in shaping sustainable development. The analysis covers European countries for the period 2014–2022. The European countries under analysis are as follows: Austria, Belgium, Bulgaria, Czechia, Cyprus, Croatia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, and Sweden. The chosen period (2014–2022) is justified by the availability of official databases of the World Bank (2026) and the European Commission (2026). The choice of a dynamic model, namely the Dynamic Panel Model (Arellano–Bond/GMM), was agreed upon because we also aim to test for temporal dependence. The choice of the GMM methodology for estimating dynamic models is justified by the specific nature of the data and the theoretical relationships investigated. The presence of the lagged dependent variable in the sustainable development modeling equation inherently generates endogeneity and leads to inconsistent estimators under OLS, Fixed Effects, or Random Effects (Nickell bias). In addition, some explanatory variables, such as digitalization, environmental taxation, and their interactions, are susceptible to simultaneous endogeneity, bidirectional causality, and the omission of unobserved variables. Since the data set is characterized by a relatively small number of periods (panel with small (t) and moderate (n)), estimation by GMM in the system (Arellano–Bover/Blundell–Bond) is the optimal approach, as it exploits internal tools obtained from the lagged values of the variables, providing consistent and efficient estimators. At the same time, the methodology allows for the control of heteroscedasticity and intragroup autocorrelation, and the validity of the instruments is rigorously verified through Hansen, Difference-in-Hansen, and Arellano–Bond tests for serial autocorrelation.

The general formula for the dynamic panel-data System General Method of Moments (GMM) model is as follows:

$$y_{it}={\alpha y}_{i,t-1}+{{\beta }^{\prime }X}_{it}+{\gamma }^{\prime }(X_{it}+Z_{it})+{\mu }_i+{\epsilon }_{it}$$

where

• $y_{it}$ = dependent variable for country ($i=\overline{1,29}$); t refers to the year ($t=\overline{1,9}$);
• $y_{i,t-1}$ = dependent lag (captures dynamic/persistence effect)
• $X_{it}$ = possibly endogenous explanatory variables (exp_dec, rev_dec, internet, green_bonds)
• $X_{it}+Z_{it}$ = interactions (exp_internet, rev_internet) because they capture interconnected and conditional effects
• ${\mu }_i$ = country-specific fixed effect
• ${\epsilon }_{it}$ = idiosyncratic error

The interconnectedness of the roles of fiscal decentralization, green finance, and the digital economy on sustainable economic development led to the construction of the dependent variable (y) in the model through a composite index (sustain_index), which combines GDP per capita (gdp_pc), greenhouse gases (greenhouse), share of renewable energy in gross final energy consumption by sector (renew_share), and total environmental taxes (env_tax_share), normalizing each variable between 0 and 1. In the index, greenhouse gas emissions are subtracted because higher levels are harmful to sustainability.

The direct variables, namely expenditure decentralization (exp_dec), revenue decentralization (rev_dec), green bond issuance by corporates and governments (green_bonds), gross domestic expenditure on R&D (rd_exp), gross capital formation (gcf), inflation (infl), tertiary school enrollment (edu), trade (trade), show the effects of the policies taken individually. The interactions among the variables (exp_internet, rev_internet, exp_green, rev_green, internet_green) capture conditional effects and synergies. The variables included in the analysis are described in

Table 1. Variable definitions.

Variable	Definition	Data Source
gdp_pc	GDP per capita (current US$)	World Bank (2026)
exp_dec	Expenditure decentralization as share of total local expenditure in general expenditure	European Commission (2026)
rev_dec	Revenue Decentralization as share of total local revenue in general revenue	European Commission (2026)
internet	Internat penetration rate as Individuals using the Internet (% of population)	World Bank (2026)
ict_imp	ICT goods imports (% total goods imports)	World Bank (2026)
ict_exp	ICT service exports (BoP, current US$)	World Bank (2026)
green_bonds	Green bond issuance by corporates and governments	European Commission (2026)
renew_share	Share of renewable energy in gross final energy consumption by sector	European Commission (2026)
env_tax_share	Total environmental taxes as Share of environmental taxes in total tax revenues	European Commission (2026)
rd_exp	Gross domestic expenditure on R&D by sector	European Commission (2026)
gcf	Gross capital formation (current US$)	World Bank (2026)
infl	Inflation, consumer prices (annual %)	World Bank (2026)
edu	School enrollment, tertiary (% gross)	World Bank (2026)
trade	Trade (% of GDP)	World Bank (2026)
greenhouse	Greenhouse gases (CO2, N2O in CO2 equivalent, CH4 in CO2 equivalent, HFC in CO2 equivalent, PFC in CO2 equivalent, SF6 in CO2 equivalent, NF3 in CO2 equivalent)	European Commission (2026)

Source: computed by the author.

.

4. Results

For empirical evidence, the dynamic panel-data System GMM model was chosen. The results of these tests confirm the adequacy of the instruments and the correct specification of the models, thereby enhancing the robustness of the empirical conclusions. The results are presented in

Table 2. The results of the regression analysis.

	M(1) Sustain_Index	M(2) Sustain_Index	M(3) Sustain_Index
L.sustain_index	0.836 *** (0.098)	0.894 *** (0.115)	0.809 *** (0.099)
exp_dec	−0.027 (0.018)	−0.256 ** (0.108)	−0.292 ** (0.115)
rev_dec	0.025 (0.017)	0.249 ** (0.118)	0.284 ** (0.119)
internet	0.016 ** (0.006)	0.014 (0.016)	0.015 (0.014)
ict_exp	−0.000 (0.000)	−0.000 (0.000)	−0.000 (0.000)
ict_imp	−0.004 (0.007)	−0.001 (0.013)	−0.002 (0.013)
green_bonds	0.009 * (0.004)	0.068 (0.073)	0.010 (0.074)
rd_exp	−0.088 (0.054)	0.037 (0.061)	0.022 (0.069)
gcf	0.000 (0.000)	0.000 (0.000)	0.000 (0.000)
infl	0.016 (0.016)	0.016 (0.012)	0.016 (0.011)
edu	0.001 (0.002)	−0.003 (0.003)	−0.002 (0.003)
trade	−0.001 (0.001)	−0.001 (0.001)	−0.001 (0.001)
exp_green		0.003 (0.005)	0.001 (0.007)
rev_green		−0.003 (0.006)	−0.000 (0.007)
internet_green		−0.001 (0.001)	−0.000 (0.001)
internet_envtax		0.001 *** (0.000)	0.001 *** (0.000)
exp_internet		0.003 ** (0.001)	0.003 ** (0.001)
rev_internet		−0.003 ** (0.001)	−0.003 ** (0.001)
_cons	−0.856 * (0.467)	−1.222 (1.473)	−1.198 (1.319)
Observations	228	228	228

Standard errors in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01. Source: computed by the author using Stata 14.2.

.

In the M(1) model, the dependent lag (L1.sustain_index) has a coefficient of 0.836, indicating that sustainable development is highly persistent and that about 84% of the current index level is explained by its previous level. From an economic point of view, policies and economic conditions from the previous period are strongly transmitted to the current period, which suggests a cumulative effect. Expenditure decentralization has a negative but statistically insignificant effect, suggesting that integration with digitalization or other local policies may be necessary for effectiveness. Revenue decentralization have a positive but insignificant effect. We infer that fiscal autonomy can support sustainable development, but this is not strongly demonstrated here. Internet penetration has a positive, significant impact, demonstrating that digitalization contributes to sustainability. Almost significantly positive is green bonds, which shows that green financing supports sustainable development. ICT imports and ICT service exports do not significantly influence the index. The other control variables (rd_exp, gcf, inflation, education, trade) have no significant effect in this model.

In line with M(1), sustainable development is highly persistent over time, with the previous level of the index accounting for most of the current variation. Digitalization (internet) and green financing (green_bonds) contribute positively to sustainable development. Fiscal decentralization has no significant direct effect in this simple model. The model is valid, according to the AR and Hansen tests; the GMM instruments are adequate. The variables in the composite index do not appear as independent, which avoids problems of mechanical endogeneity. Hypothesis 1 is partly supported in this model, but is confirmed when considering the other two interaction models.

In the M(2) model, the Dependent Lag (L1.sustain_index) has a coefficient of 0.894, indicating that sustainable development is very persistent, approximately 89% of the current level of the index depends on the previous level. As an economic implication, previous conditions and policies continue to strongly influence sustainable development, confirming the cumulative effect. Decentralized spending negatively affects sustainable development, with a statistically significant negative effect, which may indicate inefficient allocation of the expenditure at the local level. Decentralized revenues have a positive and significant effect, which supports that fiscal autonomy stimulates sustainable development. Digitalization has a positive but insignificant effect, which in itself does not explain much of the variation without interactions. The interactions between expenditure decentralization and digitalization have a positive and significant sign, demonstrating that digitalization amplifies the effect of local spending on sustainability. Negative interaction between revenue decentralization and digitalization, demonstrating possible substitution effect or inefficient allocation of digital resources at the local level.

Significant positive interaction between digitalization and environmental taxes, demonstrating that digital policies can increase the impact of environmental taxes on sustainable development.

The other interactions and control variables (ict_exp, ict_imp, rd_exp, gcf, infl, edu, trade, exp_green, rev_green, internet_green) have small and insignificant coefficients, having a marginal contribution in this model.

According to the M(2) model, sustainable development is strongly persistent over time. Decentralized spending has a direct negative effect, but the interaction with digitalization becomes positive; digitalization can transform the effect of local spending into a beneficial one. Revenue decentralization support sustainability, but their combined effect with digitalization can be negative, indicating a possible inefficiency or allocation conflict. The interaction between digitalization and environmental taxes is significantly positive, with digital economy amplifying the effect of environmental policies on sustainable development. The remaining control variables and green financing (green_bonds) are not significant in this model.

The model is valid, according to AR and Hansen tests, and the GMM tools are appropriate.

In the M(3) model, a System GMM model with lags 3–4 and interactions, the dependent lag (L1.sustain_index) has a coefficient of 0.81, indicating strong inertia in sustainable development. Even with lags of 3–4 as instruments, this effect remains robust. Expenditure decentralization alone reduces sustainable development, perhaps due to an inefficient allocation at the local level. Revenue decentralization stimulates sustainable development, showing the benefit of fiscal autonomy. Simple digitalization does not directly influence the index. Green finance has no clear direct effect in this model, but may be relevant through interactions. The interaction of decentralized spending and digitalization shows that digitalization amplifies the positive effect of local spending. The interaction of decentralized revenues and digitalization may indicate a less efficient allocation of digital resources. Digital economy amplifies the effect of environmental taxes on sustainable development.

Therefore, the M(3) model shows the persistence of sustainable development, which is very high, suggesting that previous policies and conditions strongly affect the current level. Fiscal decentralization expressed by expenditure decentralization alone has a negative effect, but the interaction with digitalization (exp_internet) turns it into a positive one; and decentralized revenues are beneficial, but the interaction with digitalization (rev_internet) can reduce efficiency.

The interaction between the digital economy and environmental taxes has a significant positive effect, justifying that technology enhances the impact of environmental policies on sustainability. The other variables do not show significant direct effects, but may influence indirectly through interactions or specific contexts.

The model confirms the combined role of fiscal decentralization, digital economy, and green finance in sustainable development, with direct and interactive effects. Lags 3–4 for instrumentation preserve the robustness and validity of the estimates.

If we compare the three models, we find that all are robust, free of second-order autocorrelation, and use valid instruments (

Table 3. Comparison approach between models.

Characteristics	Model 1	Model 2 (Interactions, Lag 2–3)	Model 3 (Interactions, Lag 3–4)
Dependent	sustainable_index	sustainable_index	sustainable_index
Independent variables	Expenditure/revenue decentralization, Internet, ICT, green bonds, R&D, GCF, infl, edu, trade	The same variables + interactions: exp_green, rev_green, internet_green, internet_envtax, exp_internet, rev_internet	Same as model 2, but instrumentation lag 3–4
Lag dependent	L1	L1	L1
No of instruments	16	21	21
F-statistic	216.59	904.94	153.41
AR(2) test	p = 0.785	p = 0.738	p = 0.768
Hansen test	p = 0.184	p = 0.553	p = 0.328
Sargan test	p = 0.198	p = 0.553	p = 0.258

Source: computed by the author.

).

Sustainable development (

Table 4. The persistence of sustainable development.

Model	Coefficient	Interpretation
1	0.821	Sustainable development is persistent; strong effect of the previous level.
2	0.894	Inertia increases slightly when we include interactions; the past becomes more determining.
3	0.809	Lag 3–4 slightly reduces the coefficient, but it persists strongly.

Source: computed by the author.

) depends stably on previous values, regardless of interactions or lags, identifying the persistence of sustainable development (L1.sustain_index).

From the perspective of fiscal decentralization, expenditure decentralization becomes significantly negative in the models with interactions, and revenue decentralization is positively significant. The interactions partially explain the change in the effect, because digitalization and green finance modulate the impact of decentralization.

From the perspective of interactions, the model M(1) model has no interactions. Models M(2) and M(3) include exp_internet and rev_internet, with significant coefficients, with opposite effects: expenditure decentralization and internet increase sustainability; revenue decentralization and internet decrease efficiency. The interaction internet_envtax is positive and significant in models M(2) and M(3), meaning that digitalization amplifies the impact of environmental taxes. Direct variables such as internet, green_bonds, rd_exp, or gcf are not significant in any model. Larger lags (3–4) preserve the robustness of the instruments (Hansen/Sargan).

The interactions explain more of the index variation and show how digitalization can modulate the effects of decentralization and environmental policies.

In conclusion, model M(1) is simple and of limited significance, without interactions, showing only the persistence of sustain_index and fiscal decentralization. In model M(2), by adding interactions (lag 2–3), it reveals the complementary effects between the digital economy and environmental taxes, plus the modulating impacts of decentralization. In model M(3) with Lag 3–4, it confirms the results of model M(2), preserving the validity and robustness; the coefficients are comparable, but longer lags increase the reliability of the instruments and confirm hypothesis 2. According to hypothesis 2, models with larger interactions and lags are more informative for economic policy.

5. Discussion

The analysis of the three dynamic specifications estimated by System-GMM highlights several important aspects that contribute to the understanding of the determinants of sustainable development in the economies analyzed, answering the main question of whether each factor—fiscal decentralization, green finance, and the digital economy—or their interactions influence the level of sustainable development.

1. The persistence of sustainable development is consistently high in all models

The lag coefficient L.sustain_index remains strong and significant (0.80–0.84), which confirms the existence of a dynamic process in which past performances decisively influence future levels. This indicates a structural inertia, meaning that countries with high sustainability values tend to remain performing well, while those with low values face difficulties in recovery without substantial interventions, as also the literature identified (Zahra & Badeeb, 2022).

2. The simple model indicates modest effects of standard economic variables, but the extended models reveal more complex systemic relationships. In the first model (without interactions), many economic and technological variables (educ, gcf, ict_exp/ict_imp, trade) have insignificant coefficients, resulting in a reduced or diffuse direct impact. Only the internet has a positive and significant effect in model 1, suggesting a moderating role of basic digitalization. This shows that the direct effects of macro variables are not sufficient to explain the variation in sustainable development.

3. Once the interactions are introduced, it becomes clear that digitalization, environmental taxation and public spending interact structurally.

Model 3, the most complete and best specified, shows that the digitalization–environmental taxation interactions (internet_envtax) have a strong and significant positive effect, with digitalization increasing the efficiency of environmental policies. The literature supports the idea of a coordination relationship between the digital economy and the ecological environment, and empirically tests that the digital economy can improve environmental quality (Li et al., 2021).

The digitalization–expenditure decentralization interactions (exp_internet) have a positive and significant effect, with complementary investments between government and the digital sector improving sustainability. The digitalization–public revenue interactions (rev_internet) produce a negative effect, suggesting potential budgetary tensions or substitution effects (digital revenues may reduce traditional fiscal resources). The literature identifies the synergistic effects of fiscal decentralization and the digital economy (Z. Liu et al., 2024).

These results indicate that variables in isolation do not determine sustainable development, but by the way economic policies, digitalization and taxation interact.

4. The significance of fiscal variables is strengthened in the extended model

Public expenditure (exp_dec) becomes negative and significant in model 3, suggesting potential inefficiencies of traditional public expenditure when not oriented towards sustainability. Public revenue (rev_dec) becomes positive and significant in the extended model, suggesting a positive relationship between fiscal capacity and sustainable development.

5. Diagnostic tests confirm the validity of the models

In all three models: they pass the Hansen/Sargan tests, so the instruments are valid; they pass the AR(2) test, so there is no second-order autocorrelation in the differences, so the dynamics are correctly modeled; they have a controlled number of instruments, so they avoid over-identification. Model 3 remains statistically the most robust.

The evolution from the simple model to the interaction model shows that determining sustainable development is a complex and interdependent process, in which digitalization, fiscal policy, and public investment reinforce or counteract each other. Simple models underestimate the strength of these links, whereas the complete model captures the underlying mechanisms by which economic policies influence the sustainable transition. In this context, both hypotheses were validated.

6. Conclusions

This study has successfully answered the research question by examining the interconnected roles of fiscal decentralization, green finance, and the digital economy in shaping sustainable development. The analysis of the three GMM models highlights that sustainable development, seen as an index, is a deeply inertial and structurally determined process, in which past performance strongly influences future developments, namely that countries with high sustainability values tend to remain performing, and those with low values face difficulties in recovery without substantial interventions.

The first model, with simple specifications that include only the direct effects of economic and technological variables, only partially captures the generating mechanisms of sustainable development. Although some traditional determinants show stable effects (revenue and expenditure decentralization), most standard economic variables do not show significant direct influences on sustainable development, as measured by an index, suggesting a much more complex relationship than anticipated by conventional models. The introduction of interactions in the models M(2) and M(3) reveals the impact of digitalization, environmental taxation and public policies through complementary mechanisms. The extended models capture the integrated nature of the sustainable transition, in which technological development, administrative capacity, and fiscal policies form an interdependent ecosystem.

From a public policy perspective, financial decentralization influences development, but it must be combined with efforts to identify common sources of public revenue and alternative revenue-generating options. From a local public expenditure standpoint, it is essential to justify and properly implement these measures in accordance with the principles of economy, efficiency, and effectiveness. Policymakers should also consider that the effectiveness of decentralization depends on both institutional maturity and environmental accountability mechanisms. Green finance policies across Europe are shaped by the European Green Deal, which uses the EU taxonomy to provide a shared definition of environmentally sustainable economic activities for both financial and non-financial companies. The Digital Single Market is highlighted by the European Policy Centre (2025) as a priority for the European Commission, which could increase the EU’s GDP by more than 4%.

In conclusion, sustainable development can be explained partly by isolated economic factors, but the interaction between digitalization, fiscal policies, institutional behaviors, and public investments creates an interdependent ecosystem. In this framework, policies that combine digital tools, environmental taxes, and strategically targeted public investments can produce a synergistic effect that accelerates the transition toward sustainability. Therefore, the final model offers not only a solid empirical explanation, but also an important conceptual framework for designing coherent and effective public policies.

The study has some weaknesses that need consideration. Since it relies on quantitative indicators, the main limitation is the availability of data for only 29 European countries from 2014 to 2022. The chosen timeframe (2014–2022) is based on the availability of official databases from the World Bank (2026) and the European Commission (2026).

Future research directions include expanding the analysis to categorize countries as developed and developing worldwide. To improve the validation of results, given the diverse economic and cultural features of countries worldwide, it would be helpful to include additional control variables in the analysis.