Do Environmental Taxes Stimulate Eco-Investments? Evidence from Seven EU Member States and the EU-27

1. Introduction

The European Green Deal sets the ambitious goal of the EU becoming climate-neutral by 2050. This transition requires massive capital investments in clean technologies and environmental protection - but also new financial risks. Companies face the so-called transition risk - the financial risk associated with the changing regulations, tax rates, technological standards, and market preferences that accompany the ecological transition (TCFD, 2017; ECB, 2020). One of the key instruments of this transition is environmental taxation.

Environmental taxes - taxes on energy, transport, pollution, and resources - are a fiscal instrument based on Pigouvian logic. By raising the price of pollution, they are intended to incentivise enterprises to invest in cleaner technologies (Pigou, 1920; Baumol, 1972). The logic is compelling. But does it work in practice?

The existing literature provides an ambiguous answer. Some studies find a positive relationship between environmental taxation and environmental outcomes, such as reduced emissions (Andersson, 2019; Sen & Vollebergh, 2018) or eco-innovations (Acemoglu et al., 2012). Others, however, indicate that the effect depends strongly on the level of the tax rate, the institutional environment, and sectoral specificity (Dechezleprêtre & Sato, 2017; Marin & Vona, 2021). Few studies employ sectoral panel data to examine the direct relationship between environmental taxation and the capital eco-investments of enterprises. Most studies use aggregated national data, which does not allow sectoral differences to be accounted for - yet it is precisely these differences that are of key importance for the management of financial risk from the ecological transition.

This study fills that gap. Using a country-sector panel of 7 countries and 4 sectors under NACE Rev.2 for the period 2014-2023, the following research question is posed: Does a statistically significant relationship exist between environmental taxes and capital eco-investments at the sectoral level in the EU? The answer to this question is important not only for environmental policy, but also for financial management - if taxes do not stimulate investments, they represent a pure fiscal burden without a corresponding environmental effect, which is a risk both for enterprises and for public finances.

2. Literature Review

The theoretical framework of environmental taxation is grounded in the concept of Pigou (1920). According to it, taxes on pollution correct negative externalities by bringing private costs into alignment with social ones. Baumol (1972) formalises this idea by proposing the so-called price-and-standard tax - a practical alternative to the theoretically optimal Pigouvian tax. The logic is straightforward - the higher price of pollution incentivises firms to invest in cleaner technologies, simultaneously reducing emissions and long-term costs.

The Porter Hypothesis (Porter & van der Linde, 1995) extends this argument. According to it, well-designed environmental regulations can not only stimulate innovation but also enhance the competitiveness of firms. The double dividend hypothesis (Goulder, 1995) adds a further argument - revenues from environmental taxes can replace distortionary taxes (e.g. income taxes), generating simultaneous environmental and economic benefit.

This outcome, however, is not automatic. Bovenberg & de Mooij (1994) emphasise that in a real environment with multiple taxes and market distortions, the net effect may be weaker than expected. Köppl & Schratzenstaller (2023) and Zhang et al. (2024) confirm that results vary substantially according to exactly how the tax is structured, how the regulation is measured, and in what institutional environment it is applied.

Thus far, the theory outlines why environmental taxes ought to work. The empirical literature, however, presents a more complex picture. The effects operate through different channels and are not always unambiguous.

The first and most studied channel of action of environmental taxes is their impact on greenhouse gas emissions and pollution. Andersson (2019) conducts one of the first quasi-experimental studies, analysing the carbon tax in Sweden through the synthetic control method. The result is that transport CO₂ emissions decrease by approximately 11% following the introduction of the tax. Sen & Vollebergh (2018) apply instrumental variables and establish that an increase of 1 euro in energy taxes leads to a reduction in carbon emissions of 0.73% in the long run. Wolde-Rufael & Mulat-weldemeskel (2023) confirm these results with panel cointegration tests for 20 European countries, finding a significant negative relationship between environmental taxes (total, energy, and transport) and CO₂ emissions. Bretschger & Grieg (2024) demonstrate for the United Kingdom that higher fuel taxation leads to a reduction in transport CO₂ emissions, without a clearly expressed negative effect on economic activity. Kohlscheen et al. (2025) establish that carbon pricing is associated with a statistically significant reduction in CO₂ emissions, particularly in the long run.

Not all studies, however, confirm this picture. Morley (2012) using GMM panel estimates for 25 European countries (1995-2005), establishes an effect on pollution but not on energy consumption. This suggests that the reduction in emissions stems from a technological transition rather than from lower consumption. Liobikienė et al. (2019) analyse 28 EU member states (1995-2012) and find no significant effect - neither direct nor indirect - of energy taxes on greenhouse emissions. The authors recommend reforming tax policy and combining it with the emissions trading system (ETS). Silajdžić & Mehić (2018) on the basis of data for 10 Central and Eastern European countries, find no convincing evidence that environmental taxes lead to effective restriction of carbon dioxide emissions. Godawska (2024) also reports heterogeneous results, establishing no statistically significant effect on nitrogen oxide emissions and only a weak effect on sulphur oxide emissions.

This group of studies reveals an important pattern. Environmental taxes may have an effect in some contexts yet appear weak or insignificant in others. The differences arise from the design of the tax, the time horizon, the method, and the chosen indicator. Therefore, the key question is not simply whether taxes work, but what exactly is being measured as an effect.

The second key channel is the relationship between environmental taxation and ecological innovations and investments. Acemoglu et al. (2012) develop a model of directed technical change. The model demonstrates that taxes and subsidies can redirect innovative efforts from dirty towards clean technologies. Aghion, et al. (2016) confirm this empirically in the automotive industry - higher fuel prices stimulate the patenting of electric and hybrid vehicles. Further studies confirm this relationship, demonstrating that environmental policies and price incentives encourage the development of cleaner technologies and increase low-carbon innovations in companies ((Johnstone et al., 2010; Calel & Dechezleprêtre, 2016; Feng et al., 2024; Hu et al., 2025).

This strand of the literature is important because it demonstrates that price pressure can alter the direction of technological development. However, it does not provide an automatic answer to the question of whether innovations translate into real eco-investments at the sectoral level. The difference between innovation (patents, R&D) and capital execution (actual investment expenditure) is substantial.

Popp (2002) and Jaffe & Palmer (1997) demonstrate that higher energy prices stimulate patenting and R&D expenditure, but evidence for a real investment effect is weaker - firms respond first with preparation rather than with capital outlays. Benatti, et al. (2024) confirm this - environmental regulations stimulate clean innovations, but the effect is considerably stronger in combination with direct R&D subsidies.

The literature on renewable energy adds a further important qualification. Johnstone et al., (2010) demonstrate that renewable energy policies influence patent activity, but results depend on the type of technology and the type of instrument. It is not sufficient to speak in general terms about environmental policy - what matters is the precise nature of the instrument: tax, subsidy, standard, or combination (Palage et al., 2019; Hille et al., 2020; Blagoeva & Georgieva, 2023).

Dechezleprêtre & Sato, (2017) present a systematic review of the literature on the impact of environmental regulations on competitiveness. The conclusion is nuanced - regulations lead to significant but small negative effects on trade, employment, and productivity in the short run - principally in energy- and pollution-intensive sectors. At the same time, there is evidence of stimulating innovation, but the benefits are insufficient to compensate the costs for regulated enterprises. Ambec et al. (2013) summarise the literature on the Porter Hypothesis and emphasise that positive effects are more likely when policies are well-designed, predictable, and consistent. Marin & Vona (2021), analysing micro-data from French manufacturing enterprises, establish that the impact of energy prices depends critically on sectoral specificity and firm characteristics. Calel & Dechezleprêtre (2016) demonstrate that the EU ETS stimulates low-carbon innovations among regulated firms. This result is an important counterpoint - price instruments can work, but the effect depends on the specific mechanism, scope, and institutional environment.

Of particular importance for the present study is the research of Carfora et al., (2021). The authors analyse the role of environmental taxes and public support policies for investments in renewable energy in the EU. Using spatial econometric models for 26 member states (2007-2018), they establish that the tax burden acts more as a barrier to renewable energy investments than as an incentive. At the same time, targeted support policies - including feed-in tariffs - demonstrate a positive effect. This result draws an important distinction - tax pressure and investment incentive are not the same thing. A sector may bear a high tax burden without this automatically leading to greater ecological investments. It is precisely this distinction that is also key to our analysis, which examines not the general ecological effect but specifically the investment response. The conclusion of Carfora et al. (2021) is in accordance with the broader literature. The effect of price instruments depends on accompanying policies. When taxes are combined with predictable support, access to financing, and clear regulatory signals, the likelihood of an investment response is higher. When such a package is absent, the tax is perceived primarily as a cost.

The growing awareness of the financial risks associated with climate change has led to the creation of institutional frameworks for the management of transition risk. TCFD (2017) defines four principal categories of climate financial risks - regulatory, technological, market, and reputational. Environmental taxes fall precisely within the category of regulatory risks, since changes in tax rates or in the scope of taxation can alter the cost structure of enterprises, their profitability, and their investment decisions. The question of whether environmental taxes stimulate eco-investments is therefore not only ecological but also financial-managerial. In this connection, ECB (2020) formulates expectations for banks to integrate climate risks into management and disclosure, whilst Nerlich, et al. (2025) identify a significant gap between the required and the actually realised green investments in the European Union. This supports the conclusion that tax policy in itself is insufficient if it is not combined with capital market reforms and regulatory predictability (Krastev & Krasteva-Hristova, 2024). The connection with the present study is direct - when taxes increase cost pressure but do not lead to investment adaptation, transition risk rises both for enterprises and for their creditors. Battiston, et al. (2017) demonstrate that such effects can be transmitted through the financial system and, through network effects, create systemic risk.

This perspective is important for the interpretation of our results. Even when no direct relationship between taxes and eco-investments is observed, this does not mean that tax policy is insignificant. Rather, it means that the effect may manifest through other channels - cost pressure, changes in the risk profile, technological deferral, or the redirection of investments. It is precisely for this reason that the empirical verification of the direct relationship is necessary rather than superfluous.

In this sense, the literature on climate financial risk and the literature on environmental taxation meet in a common question - not simply whether the policy is “green,” but whether it is effective as a mechanism for real adaptation. The present study rests precisely upon this intersection - between fiscal instrument, investment behaviour, and risk management under conditions of green transition.

The literature review reveals three substantial gaps. First, most studies focus on emissions rather than on capital eco-investments - the tax-investment relationship remains poorly studied, despite its key significance for the assessment of fiscal risk. Second, few studies combine the cross-country and cross-sector dimension, even though sectoral differences are fundamental - the same tax may act as an incentive in one sector and as a barrier in another. Third, the predominant use of absolute values without eliminating the influence of the size of the economy can lead to systematically erroneous conclusions.

The present study addresses these gaps with a country-sector panel of 7 countries and 4 NACE sectors for the period 2014-2023. Unlike most studies, which work with aggregated national data in absolute terms, here we combine the cross-country and cross-sector dimension with relative values and a five-step empirical strategy - descriptive statistics, correlations with relative values, fixed-effects panel regressions, the Granger test, and seven robustness checks. The focus is on the effectiveness of environmental taxes as a financial instrument for the management of transition risk.

3. Materials and Methods

3.1. Data and Sample

The study uses panel data extracted from the publicly available databases of Eurostat. The sample covers seven EU member states (Belgium, Bulgaria, France, Hungary, Italy, Poland, and Romania), plus EU-27 as a reference aggregate. The countries are selected so as to ensure geographical and economic diversity. Both old and new EU members are included, as well as large and small economies, and Western and Eastern European countries.

The panel has two dimensions. The first is the country-sector dimension - each unit of observation is the combination of country and sector (e.g. Bulgaria - Manufacturing). The second is the time dimension: the period 2014-2023 (10 years). Four sectors under NACE Rev.2 are covered: A (Agriculture, Forestry and Fishing), B (Mining and Quarrying), C (Manufacturing), and D (Energy). This yields a balanced panel of 7 × 4 × 10 = 280 observations (320 with EU-27).

3.2. Variables

Table 1 describes the variables employed. The dependent variable is eco-investments (EcoInvest) - capital expenditure on environmental protection, reported under the EPEA (Environmental Protection Expenditure Accounts) methodology. The main independent variable is environmental taxes (EnvTax), reported in accordance with Regulation 691/2011 of the EP and of the Council (European Parliament and Council of the European Union, 2011).

In order to eliminate the effect of the size of the economy, relative indicators are also employed - eco-investments and environmental taxes as a percentage of gross value added (% of GVA) of the respective sector. All monetary values are in millions of euros, at current prices.

3.3. Analytical Framework

The analysis follows a five-step strategy (Table 2).

In the correlation analysis, the Pearson coefficient (r) is employed for the assessment of the linear relationship between variables, both in absolute values and in relative ones (as a percentage of GVA).

As the main econometric instrument, a fixed-effects (FE) panel regression by units is applied. This model absorbs all unobserved and time-invariant characteristics of the units (e.g. regulatory tradition, technological structure, geographical location) and identifies the effect from within variation. Four specifications are employed:

Мoдел 1 (log-log): ln(EcoInvestᵢₜ) = αᵢ + β1·ln(EnvTaxᵢₜ) + β2·ln(VAᵢₜ) + εᵢₜ

(1)

Мoдел 2 (intensity): InvIntᵢₜ = αᵢ + β1·TaxIntᵢₜ + εᵢₜ

(2)

Мoдел 3 (FE + TE + lag): ln(EcoInvestᵢₜ) = αᵢ + γₜ + β1·ln(EnvTaxᵢ,ₜ₋₁) + β2·ln(VAᵢₜ) + εᵢₜ

(3)

Мoдел 4 (intensity + lag): InvIntᵢₜ = αᵢ + β1·TaxIntᵢ,ₜ₋₁ + εᵢₜ

(4)

where i denotes the unit (country-sector), t denotes the year, αᵢ denotes the unit fixed effect, and γₜ denotes the time effect. All models employ clustered standard errors (SE) by units for the correction of heteroscedasticity and autocorrelation within units. The choice between FE and RE is verified using the Hausman test (Hausman, 1978).

To verify the direction of the relationship, a panel Granger causality test (Granger, 1969) is applied. Two directions are examined - whether past values of environmental taxes improve the forecast for eco-investments (taxes → investments) and the reverse (investments → taxes). Owing to the limited time dimension (T = 10), 1 and 2 lags are employed.

4. Results

4.1. Descriptive Statistics and Trend Analysis

Table 3 demonstrates that the observations in the panel are highly heterogeneous and are appreciably influenced by scale effects. The mean eco-investment per country-sector pair is 666.76 million EUR, yet the typical value is considerably lower - the median is 84.25 million EUR. This divergence means that several very large observations (including the EU-27 aggregate and large economies such as France and Italy) dominate the distribution and elevate the mean value. Consequently, mean values should be interpreted with caution, as they do not describe the “centre” of the data but are sensitive to extreme observations.

A similar picture is observed for environmental taxes: the mean value is 2,416.49 million EUR against a median of 363.70 million EUR, which again suggests a concentration of volumes in a limited number of observations and the presence of pronounced right skewness.

The relative indicators provide a more comparable perspective, as they reduce the influence of size. Expressed in this form, mean eco-investments amount to 1.40% of sectoral GVA, whilst environmental taxes stand at 4.04% of GVA. This ratio demonstrates that within the panel the tax burden (as a relative share) systematically exceeds the observed investment response, which motivates subsequent analysis of whether higher environmental taxes are associated with higher investment intensity.

Table 4 delineates substantial cross-country differences. France stands out as the largest investor in absolute terms (mean 1,056.40 million EUR), followed by Poland (268.68 million EUR) and Italy (226.66 million EUR). In relative terms, however, Belgium registers the highest investment intensity - 2.57% of GVA, followed by France (1.96%) and Romania (1.59%). Hungary records the lowest intensity (0.58%), alongside the EU-27 aggregate (1.03%).

Particularly telling is the growth dynamics. Whilst environmental taxes increase in all countries with the exception of Italy (CAGR: −0.4%), with especially rapid growth in Bulgaria (+17.9%), Poland (+12.0%), and Hungary (+9.3%), eco-investments contract in four of the seven national economies. Romania (−9.2%), Italy (−8.3%), and Hungary (−5.4%) record the steepest declines, whilst Belgium (+5.9%) and France (+4.2%) demonstrate sustained growth. This asymmetry between the rising tax burden and declining investments in a number of countries raises initial questions regarding the effectiveness of environmental taxation as an investment incentive and necessitates thorough econometric investigation.

Figure 1 illustrates the temporal evolution of eco-investments. Among the national economies (panel a), France demonstrates a clear upward trajectory - from 3,695 million EUR in 2014 to 5,371 million EUR in 2022. Belgium follows a similar positive trend. Conversely, Italy registers a sharp decline - from 1,020 million EUR (2014) to 468 million EUR (2022-2023), whilst Poland and Romania display fluctuating patterns without a clear direction. At the EU-27 level, total eco-investments increase from 13,050 million EUR in 2014 to approximately 18,000 million EUR in 2023, representing cumulative growth of 37.9% (CAGR: 3.6%). The slight decline in 2020 most likely reflects the economic disruptions caused by the COVID-19 pandemic.

Figure 2 presents the corresponding trajectory of environmental taxes. The most remarkable characteristic is the sharp acceleration in Poland after 2020, where tax revenues nearly quadruple - from 2,069 million EUR (2014) to 7,840 million EUR (2022), driven primarily by rising energy taxation. Environmental tax revenues in Bulgaria also increase sharply, with a CAGR of 17.9% over the period. Italy, the largest collector of environmental taxes in absolute terms (mean 2,261 million EUR across sectors), remains relatively stable throughout the decade. At the EU-27 level, environmental taxes increase from 43,400 million EUR in 2014 to a peak of 72,044 million EUR in 2022, followed by a slight correction in 2023 (66,098 million EUR), yielding an overall CAGR of 4.8%.

Thus far, the data have been examined in their entirety and by country. Yet behind the aggregated figures lies the question: are environmental taxes and eco-investments distributed equally across individual sectors? The answer is an unequivocal “no,” and it is precisely this unevenness that constitutes one of the most interesting signals in the data.

The sectoral decomposition in Table 5 reveals a pronounced asymmetry between the distribution of environmental taxes and eco-investments. Manufacturing (sector C) bears 67.0% of total environmental taxes yet realises only 46.1% of eco-investments. This demonstrates that manufacturing enterprises carry a disproportionately high tax burden relative to their ecological capital expenditure. The Agriculture sector (A) presents the mirror image: it contributes only 13.6% of environmental taxes yet absorbs 32.1% of eco-investments and registers the highest investment intensity (2.71% of GVA). This pattern may reflect the presence of targeted subsidies from the CAP (e.g. greening measures under the Common Agricultural Policy), which stimulate ecological investments beyond the fiscal channel.

The energy sector (D) is distinguished by the highest tax intensity (6.54% of GVA), which is consistent with its exposure to energy taxation and emissions trading mechanisms. Its investment intensity of 1.57% of GVA, whilst moderate, may underestimate total ecological capital expenditure if the costs of compliance with the EU Emissions Trading Scheme (EU ETS) are not fully reflected in the investment data.

Figure 3 presents the sectoral structure of the two indicators for the seven national economies in aggregate over the period 2014-2023. The picture reveals a distinct imbalance. Manufacturing (sector C) bears 67.0% of all environmental taxes yet realises only 46.1% of eco-investments. In other words, this sector pays disproportionately more in taxes relative to what it invests in environmental protection. Agriculture (sector A) contributes only 13.6% of tax revenues yet absorbs 32.1% of investments - most likely owing to the targeted subsidies under the EU Common Agricultural Policy.

This imbalance raises an important question: if the tax burden in itself were to stimulate investments, one would expect sectors with higher taxes to invest more. Yet the data demonstrate precisely the opposite in the case of Manufacturing. This suggests that the tax-investment relationship is not straightforward and most likely depends on sectoral specificities - a hypothesis that will be tested formally in the regression analysis.

Figure 4 offers a cross-sectional presentation of the two key intensity indicators. Panel (a) demonstrates that the energy sector of Bulgaria bears the highest mean tax burden (14.77% of GVA), closely followed by the energy sector of Romania (14.62%) and the mining and quarrying sector of Italy (8.74%). Panel (b) demonstrates that the agriculture sector of Belgium registers the highest investment intensity (9.71% of GVA), followed by the agriculture sector of France (4.68%) and the mining and quarrying sector of Romania (4.26%). Notably, a number of country-sector pairs with a high tax burden do not exhibit a correspondingly high investment intensity (e.g. the energy sector of Romania: 14.62% taxes against 1.53% investments; the energy sector of Bulgaria: 14.77% taxes against 2.01% investments). This further underscores the necessity of a multifactorial analysis that controls for unobserved heterogeneity across countries and sectors.

The descriptive analysis delineates three key preliminary conclusions. First, eco-investments and environmental taxes differ drastically across individual countries and sectors. There is no typical behaviour - Belgium and France invest increasingly, whilst Romania and Italy contract. This diversity means that answers cannot be sought through simple mean values - a panel approach is required that accounts for what is specific to each country and sector. Second, taxes and investments move in different directions. Environmental taxes increase in almost all cases, yet eco-investments decline in four of the seven countries. If taxes were automatically to stimulate investments, one would expect the two curves to move together. Since they do not, it most likely depends on how the taxes are designed, what additional policies accompany them, and which sector they fall within. Third, the sectoral picture tells its own story. Manufacturing pays the most in taxes (67% of the total) yet does not invest proportionately more. Agriculture does the opposite - it invests intensively under a low tax burden, most likely because it has access to targeted subsidies under the EU CAP. In other words, taxes alone are insufficient - what matters is the package of instruments that underpins them.

4.2. Correlation Analysis

Before proceeding to the regression analysis, an examination is conducted of whether a statistical relationship exists between environmental taxes and eco-investments - and to what extent it is real, as opposed to being explained by the influence of the size of the economy.

At first glance, the relationship appears very strong. Figure 5a demonstrates that the correlation between eco-investments and environmental taxes in absolute terms is r = 0.926 (p < 0.001). However, this is an illusion. Both variables correlate strongly with GVA (r = 0.90 and r = 0.94) and GDP (r = 0.74 and r = 0.70). Larger economies simply have more of everything - more taxes, more investments, greater GVA. This does not mean that taxes stimulate investments - it means that both grow with the size of the economy.

How does the relationship appear once the influence of the size of the economy is removed? Figure 5b presents the correlations between the relative indicators (% of GVA), which are already adjusted for the size of the sector. The result is surprising - the correlation between tax burden and investment intensity falls to r = 0.079 and is not statistically significant (p = 0.161). In other words, when the size of the economy is accounted for and the tax burden and investment activity are compared as shares of gross value added, the apparently strong relationship breaks down.

4.3. Panel Regression Analysis

The correlation analysis demonstrated that the observed simple relationship between taxes and investments is to a significant extent the result of the scale effect. However, correlation does not control for unobserved specificities of countries and sectors - such as regulatory traditions, technological structure, or the policy environment. For this reason, the analysis proceeds to fixed-effects (FE) panel regressions, which isolate within variation - that is, they answer the question: when taxes change over time within the same country-sector unit, do investments change as well?

A balanced panel of 28 units (7 countries × 4 sectors) over 10 years (2014-2023) is employed. Table 6 presents four model specifications: Model 1 examines the logarithmic relationship (elasticity), Model 2 employs relative indicators (% of GVA), Model 3 adds time effects and a lagged tax variable, and Model 4 combines intensity with a lag. All models employ clustered standard errors by units.

The results are unambiguous: not one of the four models finds a statistically significant effect of environmental taxes on eco-investments. In Model 1, the elasticity is 0.02 (p = 0.854) - practically zero. In Model 2, the coefficient of tax intensity is 1.19, yet with a very large standard error (3.89) and p = 0.760. Model 3 examines whether taxes operate with a lag - yet the lagged value is likewise insignificant (p = 0.450). Model 4 repeats the same result for the intensities.

Particularly telling is the R² within - the indicator that measures how much of the variation within units the model explains. In all four models it is below 1%. This means that when environmental taxes increase in a given sector of a given country, this is of practically no use in predicting how investments will change.

The Hausman test (Table 7) does not reject the null hypothesis in either case (p = 0.181 and p = 1.000), indicating that the RE model is acceptable. However, this does not alter the main conclusion - the RE estimates are likewise insignificant for the tax variable (Model 1 RE: β = 0.07, p = 0.566; Model 2 RE: β = 3.32, p = 0.442).

This means that when environmental taxes rise in a given sector, this does not lead to greater eco-investments. Neither immediately, nor after one year. Neither in absolute nor in relative terms. This does not prove that taxes are useless - it is possible that they operate through indirect channels or with a longer lag. However, within the framework of this panel, the direct relationship is not confirmed.

4.4. Granger Causality Test

The regression analysis found no significant relationship from taxes to investments. However, the question remains - might the relationship run in the opposite direction, with higher investments leading to higher taxes? The Granger test examines both directions - whether past values of one variable improve the forecast for the other, beyond its own lags.

The results are categorical (Table 8). Not one of the eight tests rejects the null hypothesis. Taxes do not predict investments (highest F = 1.64, p = 0.202), yet neither do investments predict taxes (lowest p = 0.318). The result is identical at lags of 1 and 2 and in both specifications (log-log and intensity). In other words, the two variables move independently of one another over time.

4.5. Robustness Checks

The question nevertheless remains of whether the absence of an effect is attributable to the particularities of the chosen model or to the composition of the sample. The main model is therefore repeated under seven alternative specifications (Table 9).

The result is consistent: all seven checks confirm the insignificance of the tax coefficient. The closest to significance is R3 (new members only: p = 0.096), yet even there it does not cross the conventional threshold of 5%. Changing the sample, excluding the COVID years, switching to first differences, and adding time effects - none of these alters the main conclusion.

These checks reinforce the confidence that the null result is not coincidental. Within the framework of this panel, there exists no econometrically confirmed direct relationship between environmental taxes and eco-investments - in either direction.

5. Conclusions

The present study poses a simple yet important empirical question - do higher environmental taxes lead to greater eco-investments in the EU? Panel data for seven member states and four sectors (NACE Rev.2: A-D) for the period 2014-2023 are employed. The result is consistent - within the data under examination, no direct positive relationship between environmental taxes and eco-investments is found. This null effect is not the consequence of one particular model specification, but recurs across the descriptive analysis, correlation checks, and panel regressions, and remains robust under the robustness tests.

First, at first glance taxes and investments appear strongly related in nominal values (r = 0.926), yet this relationship primarily reflects the size of the economy - larger sectors simultaneously have greater tax revenues and larger investment volumes. When comparable values are compared - as a share of sectoral GVA - the correlation practically disappears (r = 0.079; p = 0.161). This is important for fiscal analysis, because conclusions based solely on nominal aggregates may overestimate the real effect of policy.

Second, fixed-effects panel regressions find no statistically significant effect of environmental taxes on eco-investments across all tested specifications (log-log, intensities, lagged tax variables, and models with time effects). The explanatory power over within variation is minimal (R² within < 1%), meaning that when the tax burden changes over time within the same country-sector unit, this is not associated with a systematic change in eco-investments within the same unit.

Third, the panel Granger causality tests show no statistically significant predictability in either direction at lags of 1-2 and in both specifications (logarithms and intensities). Within the time horizon under examination, tax changes do not precede investment changes in a manner that would permit one to speak of a clearly expressed dynamic dependence - and the reverse is likewise not observed.

Fourth, the descriptive results delineate a distinct sectoral imbalance between who “pays” and where investments are concentrated. Manufacturing generates 67% of environmental taxes yet realises 46% of eco-investments, whilst agriculture accounts for 14% of taxes and accumulates 32% of investments. The polluter-pays principle evidently does not automatically translate into investment logic at the sectoral level. In all likelihood, the investment decision is determined by the regulatory environment rather than by the tax burden as such.

From a practical standpoint, the results suggest that the current model of environmental taxation requires rethinking:

• The mere existence of a tax does not guarantee an investment response. A more effective alternative is a predictable mechanism whereby a portion of revenues is returned to the sectors that generate them - for example, through co-financing of eco-projects. This improves the predictability of financial flows and reduces investment uncertainty for enterprises.
• A uniform tax approach fails to account for the fact that sectors differ in their technological structure and investment capacity. Sectoral differentiation and/or combining the tax with targeted instruments is therefore justified in cases where tax intensity is high but the investment response is weak (for example, in the energy sector).
• The absence of a direct effect supports the argument that taxes function most effectively as part of a package rather than as a standalone incentive. The attempt to manage transition risk solely through a tax instrument appears insufficient - a coordinated combination of fiscal and regulatory measures is required.
• In order to avoid the influence of scale, the effectiveness of policies should be assessed primarily through relative indicators (such as % of GVA/GDP) and through internal changes over time, rather than through absolute volumes. This is important both for public assessments and for financial analysis and ESG risk management.

The study has several limitations that outline directions for future work. The sample is limited in terms of countries and aggregated sectors, and the time horizon (T = 10) does not permit reliable testing of longer lags. The data are in current prices, which may introduce additional distortions; the use of deflated values would yield more precise estimates. The analysis examines whether taxes directly influence eco-investments, but it is possible that the effect operates through an indirect path - for example, through eco-expenditure, technological changes (patents), or changes in energy intensity. Future studies may therefore incorporate such intermediate factors and employ dynamic panel models (e.g. GMM) in order to trace the mechanism of influence more precisely.

The study does not deny the role of environmental taxes as a price signal and instrument for limiting pollution. However, in the countries and sectors under examination, the specific expectation that higher environmental taxes lead to higher eco-investments is not confirmed, which underscores the necessity of evaluating policy instruments empirically rather than by presumption.