The Intertwined Threads of Blue Economy, Inclusive Growth, and Environmental Sustainability in Transition Economies

Abstract

This research creates the critical relationship between the blue economy, inclusive growth, and environmental sustainability in 17 transitional economies from 2000 to 2022. Using panel-corrected standard errors (PCSEs) and the Driscoll–Kraay standard error regression approach, we examine how inclusive growth significantly decreases the ecological footprint while the blue economy increases these effects through sustainable marine resource utilization and clean technologies. Focusing on countries such as Argentina, Brazil, China, India, Iran, Kenya, Malaysia, Mexico, Morocco, Pakistan, Singapore, South Africa, Saudi Arabia, and Sri Lanka, this study advances the understanding of how the blue economy fosters sustainability amidst rising consumption pressures. The findings underscore the potential of technology transfer, capacity building, regional collaboration and green finance mechanisms to unlock the blue economy’s full potential for inclusive and sustainable development, offering actionable insights for policymakers and future research directions in developing and transitional economies.

1. Introduction

In the 21st century, there is an apparent contradiction: while humanity’s desire for economic growth intensifies, the fragile balance of environmental sustainability weakens [1,2]. Amid this paradox, the blue economy concept has evolved, offering the potential for achieving economic growth as well as environmental protection, especially in the underutilized ocean stretches of the developing world [3]. Nevertheless, the actualization of this commitment necessitates us to confront intricate interconnections and ask how can we guarantee that blue economy initiatives promote comprehensive growth while mitigating their environmental impact?

The main drivers that we prioritize in the blue economy are inherent imbalances among marine resources conservation as well as economic growth. For blue economy to be sustainable, it must address the underlying conflicts [4]. To effectively address the issues, it is essential to obtain a deeper comprehension of the effects and interactions of the blue economy on the overall ecosystem. The blue economy consequences and overall environment will be better comprehended because of the interface mapping between the two areas. Waste utilization technologies, such as converting chitin to chitosan, are crucial for promoting sustainability in the blue economy. These innovations support waste reduction and enhance resource efficiency, aligning with circular economy principles. By integrating waste-to-resource technologies, the blue economy can reduce its ecological footprint while fostering growth in sectors like aquaculture and marine biotechnology, contributing to both environmental sustainability and inclusive development.

This paper delves into this intricate nexus, focusing on transition economy nations navigating the precarious path from state-controlled systems to market-driven ones. With their rich coastal resources and burgeoning aspirations, these countries offer a crucial testing ground for blue economy interventions. Yet, their unique developmental challenges warrant a nuanced understanding of how blue economy endeavors translate into environmental and societal well-being.

The concept of inclusive growth (IG) posits that economic progress must extend beyond mere GDP increases, encompassing the equitable distribution of benefits and access to opportunities for all members of society. Although economic expansion is typically associated with progress, it can lead to the uneven distribution of benefits across different social groups and strain environmental resources. An analysis of IG’s environmental consequences is necessary. Blue economy efforts have the capability to worsen prevailing differences. The dominant discourse around blue growth often overlooks the unequal allocation of advantages and possible disadvantages, leading to social injustices such as dispossession, environmental degradation, and the marginalization of certain groups [5,6,7]. The literature on the blue economy has collapsed to address vital geographical concepts like space, place, scale, and power relations that can contribute to unequal development procedures and regional differentiation [8]. The discussion of the blue economy has not given enough attention to social equity and inclusion issues, which has led to persistent social injustices and the monopolization of ocean space by powerful enterprises [9]. The rapid acceleration of the ocean economy often produces some advantages and sizable social evils for coastal communities, highlighting the need for a socially sustainable and equitable blue economy that benefits all coastal populations [10]. Furthermore, Kamah and Riti [11] provide evidence suggesting that placing equal importance on well-being and economic growth can incentivize adopting cleaner technology and conserving resources, hence fostering environmental sustainability. Therefore, comprehending the complex correlation between inclusion and ecological impact in the blue economy is crucial for charting a course toward fair and sustainable progress.

However, the association amongst economic growth and environmental deprivation is often tense. On the one hand, economic growth, even when inclusive, may lead to increased resource consumption and ecological degradation [12]. This phenomenon, known as the Environmental Kuznets Curve (EKC) hypothesis, suggests that environmental pressures primarily rise with economic development however eventually decline [13,14,15,16]. However, the empirical evidence for this inverted U-shaped relationship is contested, and achieving it requires careful policy interventions and technological advancements [17,18].

Total fishery production (TFP), a vital component of the blue economy in many transition economies, further complicates this dynamic. Balancing the economic benefits of TFP with ecological sustainability requires implementing effective fishery management practices and exploring alternative, responsible aquaculture methods [19].

This study goes beyond traditional GDP metrics and explicitly investigates how blue economy endeavors affect the well-being of all segments of coastal communities. This study further unpacks the complex interplay between these critical variables, providing nuanced insights into the sustainability of blue economy practices. Maintaining biodiversity, sustaining livelihoods, and guaranteeing the long-term health of aquatic ecosystems all depend on balancing the entire yield of fisheries.

AFF (agriculture, forestry, and fishing) are three interconnected industries that use natural resources to create fuel, food, fiber, and other products [20]. The terms “agriculture,” “forestry,” and “fishing” relate to the maintenance of forests and trees for diverse uses, the raising of cattle and crops on land, and the gathering of aquatic life from bodies of water.

AFF can promote inclusive growth by giving billions of people access to food security and nutrition, creating jobs and income for millions upon millions of people, boosting trade balance and foreign exchange earnings through exports, improving poverty alleviation and rural development, encouraging the adoption of technology and innovation, protecting traditional knowledge and cultural heritage, boosting resilience to jolts and catastrophes, and preserving the services of an ecosystem like water management, biodiversity preservation, and carbon sequestration [21]. However, if AFF is not carried out sustainably and fairly, it may impede inclusive growth [22]. According to Meckling and Allan [23], AFF can lead to pollution, greenhouse gas emissions, soil erosion, deforestation, desertification, water scarcity, and biodiversity loss. They can also cause societal issues such as marginalization, exploitation, inequality, land conflicts, and human rights violations [24].

Maintaining sustainable fishery productivity and reducing ecological footprints requires balancing conservation efforts and fishery management. Emerging controls regarding biogenic amines in fresh fish and items from fishing were covered by [25]. According to Oboh [26], encouraging species diversity among farmed fish can surge Nigeria’s aquaculture production. The author also stated that aquaculture is the fastest-growing sector of the food economy and provides around half of the fish consumed worldwide.

Moreover, Sumaila and Ebrahim [19] revised their analysis of the global level of fishery sector subsidization, estimating the breadth volume, and assessed global fishery subsidies at USD 35.4 billion in 2018, of which USD 22.2 billion are capacity-enhancing subsidies. Of the overall anticipated subsidy, the five countries contributing the most (China, EU, USA, Korea, and Japan) account for 58% (or USD 20.5 billion). Since the most current estimate in 2009, which was USD 41.4 billion in 2018 constant terms, the updated global amount has dropped. The discrepancy in the actual quantity of subsidies given and the improvements in methodology account for a considerable portion of the gap between these two estimations. Therefore, a straight statistical comparison of these figures might not be suitable.

Many academics have suggested a complicated and nuanced relationship between resource depletion and ecological footprints, as suggested by [27]. There are numerous forms of resource depletion, including the depletion of water resources, the loss of biodiversity due to deforestation, agricultural practices that degrade soil, the depletion of industrial resources and mineral resources, climate change, and the lessening of fossil fuels. The following references shed light on this relationship: “Wackernagel, Schulz [28] define ecological overshoot and explain how resource depletion and rising ecological footprints result from human resource consumption exceeding Earth’s regeneration capacity. According to Hoekstra and Mekonnen [29], water resources are crucial in ecological footprints. The depletion of freshwater resources primarily causes environmental stress due to excessive use and poor management, which impacts terrestrial and aquatic ecosystems. Deforestation significantly affects ecological footprints, which results from over-logging and agricultural growth [30]. Ecosystem disruption and biodiversity loss increase the environmental effect and highlight the correlation among ecological footprints and natural resources reduction. Unsustainable farming methods exacerbate the ecological footprint by degrading the soil [16]”.

Resource depletion significantly affects ecological footprints, including mineral extraction for industrial operations [31]. The environmental impact is exacerbated by habitat destruction, pollution and increased carbon emissions from mining, and the processing and disposal of minerals. One of the leading bases of ecological footprints is climate change, exacerbated via fossil fuel depletion [32]. Fossil fuel combustion produces greenhouse gases that impact biodiversity and change climate patterns, highlighting the adverse effects of resource depletion on the ecosystem. Innovative technology promotes resource efficiency and environmental health, reducing ecological footprints and enhancing resilience [15].

Following, He and Wang [33] used international trade and general government final consumer spending as moderating variables to examine the influence on the ecological footprint of high-emitting nations of mineral rent and resource depletion between 1973 and 2019. International trade and mineral rent have a negative association via ecological footprint. Conversely, in the nations that produce the most, a beneficial relationship exists between an ecological footprint and the depletion of natural resources and overall government final consumer spending. A summary of important and relevant studies is provided in

Table 1. Literature Review Table.

Authors	Variables	Methodology	Conclusions
Kautsky, Berg [34]	Aquaculture production	Case studies	Colombia’s coastal environment is already being used almost entirely for shrimp aquaculture. On the other hand, a tilapia farm fed by domestic, agricultural, and fishery waste depends relatively little on different ecosystems.
Zahid, Ali [35]	Inclusive growth, renewable energy	Dynamic ordinary least square (DOLS)	IG is positively associated with EF. RE negatively associated with EF.
Saqib, Usman [36]	Renewable energy, green growth	Second generation	RE and GG have two-way causality with EF.
Xia and Liu [37]	Natural resource, Inclusive growth	Methods of moment quantile regression (MMQR)	NR and IG are positively associated with EF.
Guillen, Natale [38]	Fishery production system, aquaculture production	MRIO model	The seafood consumption footprint shows how much of the world’s seafood is consumed by all countries, and growing seafood consumption across the globe fuels overfishing, which destroys marine biodiversity and ecosystems.
Cashion, Tyedmers [39]	Production of fishmeal and fish oil (FMFO), carbon footprint, marine footprint	Mix methods	FMFO inputs can significantly influence fish feeds and fed aquaculture regarding ecological sustainability.
Ke, Dai [40]	Green innovative efficiency	The Hensen threshold regression model	GIE is positively associated with EF.
Wicki and Hansen [41]	Green technology innovation	Case study	Innovations in green technology frequently result in more sustainable and clean activities. In contrast, when implementing green technologies, significant initial expenditures may be associated with research, development, and infrastructure.
Nawaz, Azam [42]	Natural resource depletion	CIPS unit root, Granger causality	Rapid economic growth promotes NRD, which puts more of a burden on the ecosystem.
Yi, Abbasi [43]	Natural resource depletion, renewable energy	ARDL and Kernel-based regularized the least square	NRD surges the environmental footprint. Further, an increase in RE and a reduction in the use of fossil fuels are helpful for a sustainable environment.
Gössling, Hansson [44]	Ecological footprints	Statistical analysis	The study describes a technique for estimating the ecological footprint of Seychelles leisure travel, emphasizing the effects of air travel on the environment. It highlights the importance of sustainable tourism practices and calls into question the place of long-distance travel in biodiversity conservation.
Chapman and Husberg [45]	Agriculture, forestry, and fishing	Case studies	The poor enforcement of regulations and lax enforcement of existing ones result in insufficient worker protection in the agricultural, forestry, and fisheries (AFF) industry. Immigration laws and past legal “exceptionalism,” which excluded certain industry-specific regulatory protections, increase the vulnerability of AFF workers.

.

This study’s novelty lies in its focus on the blue economy within transitioning economies, an area that remains underexplored despite the growing global interest in sustainable economic models. By analyzing a diverse group of 17 transitioning economies over the period from 2000 to 2022, it offers a unique perspective on the intricate relationships between blue economy activities, inclusive growth, and ecological impact. This study introduces a novel approach by employing panel data analysis to examine how factors such as population growth and trade openness influence the development of a sustainable and inclusive blue economy in these contexts. The significance of this research lies in its potential to provide actionable insights for policymakers in transitioning economies, helping them develop strategies to mitigate the ecological impact of the blue economy while fostering inclusive development and environmental sustainability.

The organization of this document is as follows: “Sections Two and Three encompass the theoretical framework and the data sources, factors, and methodologies used, respectively. The fourth and fifth parts discuss the econometric methods and results. The sixth section is about discussion. The seventh section provides a concise overview of the conclusions. Sections 8, 9, and 10 function as the policy implications, limitations, and further recommendations”.

2. Theoretical Framework

EF is an ecological footprint measured in per capita (gha) proxy for the environment. IG stands for inclusive growth measured in GDP per person employed. RE displays the percentage of total final energy used, renewable energy and the TFP, which is total fishery production displayed in metric tons. Moreover, AP is aquaculture production shown in metric tons, and AFF stands for agriculture, forestry, and fishing as a percentage of GDP. Likewise, GTI is for green technological innovation proxied by patent application residents, and NRD indicates the natural resource depletion percentage of GNI. Following the conceptual foundation, the following discussion shows how the variables relate:

EF and IG: Through its reflection on human activity’s sustainability and environmental impact, the EF can impact IG [46]. A high EF may suggest the heavy use of natural resources and waste production, both harming the environment and limiting chances for the next generations. As a result, the relationship between the EF and IG hinges on by what means that the EF is balanced and decreased inclusively and conscientiously.

EF and RE: RE affects EF [47,48]. Due to its capacity to reduce dependence on limited resources and lower greenhouse gas emissions, renewable energy sources like wind and solar can drastically reduce ecological footprints. Supporting sustainability and halting environmental deterioration are two benefits of this shift. To optimize the net ecological advantages, however, sound life cycle management is required because the development and disposal of renewable technology can harm the environment.

EF and TFP: The whole fishery production affects the ecological performance [39]. Overfishing, fish stock depletion, the disturbance of marine ecosystems, and increased ecological footprints due to habitat damage and species extinction can all be caused by uncontrolled or unsustainable total fishery productivity. By guaranteeing a balanced approach to resource extraction, on the other hand, sustainable and well-managed fishing productivity can help maintain healthy ecosystems, preserve biodiversity, and lessen ecological footprints.

EF and AP: Aquaculture output might affect ecological footprints depending on how sustainably the procedures are used [34]. Ethical and environmentally sustainable aquaculture techniques can help reduce footprints by reducing the impact on habitat and resource utilization. On the other hand, problems like the exploitation of resources and habitat deterioration can cause intensive or poorly managed aquaculture to worsen ecological footprints.

EF and AFF: Through significant land use, deforestation, and resource extraction, the combined activities of agriculture, forestry and fishing impair biodiversity and ecosystem services and add to ecological footprints [45]. Overfishing, deforestation for agricultural development, and intensive agriculture methods can all substantially impact ecological footprints by destroying natural resources, emitting greenhouse gases and upsetting ecosystems. Implementing sustainable farming, forestry, and fishing methods is necessary to minimize ecological footprints, conserve biodiversity, and maintain the long-term well-being of ecosystems.

EF and GTI: Green technology innovation, which encourages resource efficiency, greener energy sources, and sustainable activities [41], is essential to lowering ecological footprints. Using environmentally friendly technologies in various sectors promotes a more sustainable and conscientious use of resources by reducing adverse environmental effects. Achieving a more ecologically sustainable future and reducing ecological footprints requires developing and accepting green technologies.

EF and NRD: The depletion of natural resources and ecological footprints [42] are caused by human actions, including over-extraction, deforestation and excessive resource use. Thus, taking care of ecological footprints means implementing sustainable measures to lessen the continuous loss of natural resources and promote a more peaceful coexistence of human activity and the environment.

3. Description of Variables, Data Sources, and Model Specification

Data are taken from transition economies to ensure the impact of the blue economy and inclusive growth on sustainability of the environment. Out of 24 transition economies, 17 countries (Argentina, Brazil, China, India, Iran, Kenya, Malaysia, Mexico, Morocco, Pakistan, Singapore, South Africa, Saudi Arabia, Sri Lanka, Thailand, Turkey, United Arab Emirates) were chosen based on data availability. The data of these countries were taken from different sources mentioned below in

Table 2. Variable description and sources.

Variables	Symbol	Definitions	Sources
Ecological Footprint	EF	Ecological Footprint vs. Biocapacity (gha per person)	Global Footprint Network (GFN)
Green Technological Innovations	GTI	All Technologies (total patents)	OECD
Inclusive Growth	IG	GDP per person employed (constant 2017 USD)	WDI
Total Fishery Production	TFP	Total Fishery Production (metric tons)	WDI
Aquaculture Production	AP	Aquaculture Production (metric tons)	WDI
Agriculture, Forestry, and Fishing	AFF	Agriculture, Forestry, and Fishing, value added (Constant 2015 USD)	WDI
Renewable Energy Consumption	REC	Renewable Energy Consumption (% of total final energy consumption)	WDI
Natural Resource Depletion	NRD	Adjusted Savings: natural resources depletion (% of GNI)	WDI

and the analysis was performed by STATA software17. The period includes 2000–2022. The data sources comprise World Development Indicators (WDI), OECD, and the Global Footprint Network (GFN). The ecological footprint (EF) is the dependent variable, an environmental sustainability proxy. The blue economy and inclusive growth (IG) are independent variables. However, control variables are green technological innovation (GTI), renewable energy consumption (REC) and natural resource depletion (NRD). The components of blue economy consist of aquaculture production (AP); agriculture, forestry and fishing (AFF); and total fishery production (TFP) [49]. The definition of each variable, along with its data source, is given below in Table 2:

Based on the theoretical framework, the following model is developed:

$$E{F}_{\mathrm{it}}={\beta }_0+{\beta }_1{\mathrm{IG}}_{\mathrm{it}}+{\beta }_{2 }{\mathrm{AP}}_{\mathrm{it}}+{\beta }_{3 }{\mathrm{TFP}}_{\mathrm{it}}+{\beta }_{4 }{\mathrm{AFF}}_{\mathrm{it}}+{\beta }_{5 } {\mathrm{GTI}}_{\mathrm{it}}+{\beta }_{6 } {\mathrm{REC}}_{\mathrm{it}}+{\beta }_{7 }{\mathrm{NRD}}_{\mathrm{it}}+{\epsilon }_{\mathrm{it}}$$

(1)

EF indicates ecological footprint; IG shows inclusive growth; AP describes aquaculture production; TFP shows total fishery production; AFF is agriculture, forestry, and fishing; GTI is green technological innovations; REC is renewable energy consumption; and NRD is natural resource depletion. ${\beta }_0$ indicates the constant term. The coefficients of independent variables are shown by ${\beta }_1, {\beta }_{2 }, {\beta }_{3 }, {\beta }_{4 }, {\beta }_{5 }, {\beta }_{6 }$ and ${\beta }_{7 }$. $\epsilon$ is the error term. i shows countries; however, t shows the period (2000–2022). The relationship between the EF and IG is justified by the work of Raitzer and Wong [50]. The EF and AP are also linked as justified by a study conducted by Zepeda and Jones [51]. Geng and Wu [49] justified the link of the EF with the TFP and AFF. A study conducted by Usman and Hammar [52] depicts the relationship among EF and GTI. The connection between the EF and REC is found by Destek and Sinha [53] and Sun and Bao [54]. A study conducted by Ahmad and Jiang [55] describes the relation between the EF and NRD.

The above Equation (1) is converted into natural logarithmic form to restrain heteroscedasticity and the effect of outliers. As a result, it describes the elasticity coefficient among dependent and independent variables [56]. Equation (1) is then converted to Equation (2) as follows:

$${\mathrm{lnEF}}_{\mathrm{it}}={\beta }_0+ln{\beta }_1{\mathrm{IG}}_{\mathrm{it}}+ln{\beta }_{2 }{\mathrm{AP}}_{\mathrm{it}}+ln{\beta }_{3 }{\mathrm{TFP}}_{\mathrm{it}}+ln {\beta }_{4 }{\mathrm{AFF}}_{\mathrm{it}}+ln{\beta }_{5 } {\mathrm{GTI}}_{\mathrm{it}}+ln{\beta }_{6 } {\mathrm{REC}}_{\mathrm{it}}+ln{\beta }_{7 }{\mathrm{NRD}}_{\mathrm{it}}+{\epsilon }_{\mathrm{it}}$$

(2)

4. Econometric Methods

The econometric methods comprise various tests such as prerequisite diagnostic tests, unit root tests, cointegration tests, and regression estimation. First, we discuss prerequisite tests.

4.1. Prerequisite Diagnostic Tests

4.1.1. Cross-Sectional Dependence (CD) Test

This indicates that, because of globalization, cross-country connections and worldwide economic integration, an interruption in one economy undoubtedly affects other economies [57]. This results in void test statistics along with effectiveness harm. Therefore, before performing regression analysis, the panel’s CD must be confirmed [58,59]. Several CD tests such as (i) the Breusch and Pagan LM test [60], (ii) the Pesaran CD test [61], and (iii) the Pesaran, Ullah, and Yamagata bias-adjusted LM test [62] were used in this study. The null hypothesis indicated no CD in the panel [58]. The Breusch and Pagan LM test is suitable for panels with long durations (T) and small cross-sections (N).

4.1.2. Slope Heterogeneity

Slope heterogeneity across countries must be checked employing the null hypothesis (Ho: βi = β for all i) and alternative hypothesis (H₁: β_i ≠ β_j for i ≠ j) to account for country-specific characteristics [63,64]. By small N and large T, entirely exogenous explanatory variables are present, and the error variance is homoscedastic; the conventional F-test is used. The slope homogeneity test proposed by Swamy [65] may be used when heteroskedasticity is present. These tests are appropriate whenever T > N [66]. For large samples (N, T→∞), Pesaran and Yamagata [67] provide the Δ test or slope homogeneity test. Chang, Chu et al. [68] estimate the standardized dispersion statistic by computing a modified version of Swamy’s test.

4.1.3. Autocorrelation

Its consequences are bias, ineffective findings, and standard errors. A novel, simple autocorrelation test with minimal assumptions was described by Wooldridge [69]. It eliminated the influence at the individual level by using the residuals from linear regression in the first differences. The expression is [70] with constant and time-invariant variables.

4.1.4. Multicollinearity

The relationship between the independent variables in the model is indicated by the term multicollinearity [71]. Many consequences result from it, including (a) inconsistent regression parameters concerning sign or magnitude; (b) insignificant coefficients even with high correlation and R2; (c) drastic changes in regression coefficients as a result of small changes in the data; and (d) misleading conclusions because of increased confidence intervals and standard errors. The current research used regression output and a variance inflation factor (VIF) to confirm multicollinearity. According to Kalnins [72], in general, when the highest VIF value is less than 10, 8, or 5, multicollinearity is unlikely to occur.

4.1.5. Heteroscedasticity

A linear regression model assumes that the error term’s variance is constant. Its violation results in heteroscedasticity, an econometric issue. According to Alabi, Ayinde [73], it causes incorrect hypothesis testing and biased OLS estimates. A modified Wald statistic utilizing the residuals of a fixed-effect regression was employed to examine the group-wise heteroskedasticity [74].

4.2. CIPS Panel Unit Root Tests

For regression analysis, the exact integration order is a precondition. Constant mean, variance, and covariance were demonstrated using a stationary time series. On the other hand, non-stationary variables can result in outcomes, i.e., (i) the forecast turns into an insignificant (ii) problem in the selection of a suitable model and (iii) false consequences, still before a high R2, a significant t-test, and F-test [59]. Banerjee and Cockerell [75] reported poor size characteristics and over-rejection as a result of CD in the panel. Therefore, traditional unit root tests are inappropriate. The second-generation unit root test, also known as the cross-sectionally augmented IPS (CIPS) test [76], is suitable. Pesaran [76] provides the CIPS test by expanding the unit root test of Im and Pesaran [77] because of CD. The series is stationary since it rejected the null hypothesis, which is non-stationarity (bi = 0 for all i).

4.3. Westerlund Panel Cointegration Tests

To prevent misleading regression, the long-run link amongst the variables is necessary. For the analysis to be conducted, cointegration must be verified. According to Ali and Yaseen [78], the cointegration of two or more variables is implied by the long-term relationship. If the linear combination of two variables that were non-stationary independently turned stationary, then there would be the cointegration of order (1.1) between them [79]. The Westerlund [80] cointegration test, which is a second-generation cointegration test, may be used. Because of its high power and accuracy [81], it performs better than residual-based tests [82].

4.4. Regression Analysis

4.4.1. Driscoll–Kraay Standard Error Approach

Problems of econometric inconsistencies in the panel data analysis arise, particularly with more selected countries than the time period (N > T). On the other hand, heteroscedasticity with autocorrelation leads to an inflation in R². In doing so, a review of the numerous studies and the estimation using general regression techniques (pooled OLS, random effects, and fixed effects) may be misleading. The present study, therefore, uses the D/K errors from the economic growth paper written by Driscoll and Kraay [83]. The OLS coefficient estimation method refers to a non-parametric technique of estimation of panel data regression coefficients. This has several advantages, which means (a) it can handle CD and observations, (b) it is capable of managing missing observations as well as heteroscedasticity, and (c) it works with panels that are balanced or imbalanced [84]. The same approach was applied by Opoku-Mensah and Chun [85]; Khan and Idrees [86]; and Karimli and Mirzaliyev [87].

We are investigating the effect of the blue economy and inclusive growth on environmental sustainability (ecological footprint) for a panel of 17 transition nations applying the Driscoll and Kraay [83] standard error approach. We start by calculating the averages of the product of the independent variables and residuals to create standard errors with the additional aspect of being robust as opposed to cross-sectional dependence. These values are then used in a weighted HAC estimator, which is how the D/K standard error technique works [88]. When heteroscedasticity and serial and spatial dependence are possible in the data, the D/K standard error is supposed to be one of the better approaches [89,90]. The D/K method is a large-scale, non-parametric strategy that offers flexibility.

Furthermore, the D/K covariance estimator applies to both balanced and unbalanced panel data and can accommodate missing values. In general, D/K estimations exhibit robustness against both temporal and cross-sectional dependency. Based on a linear model that may be described as follows, this research employes D/K standard errors to calculate pooled Ordinary Least Squares (OLSs).

yi;t = x’_i;tβ + ε_i;t;

(3)

I = 1;…;N; t = 1;….;T

(4)

whereas the dependent variable (ecological footprint) is y_i,t, and x_i,t describes the explanatory variables (blue economy, inclusive growth, renewable energy consumption, green technological innovation, and natural resource depletion). i represents the cross-section (country), N shows the number of countries from 1 to 17, t depicts time, and T describes the number of years from 2000 to 2022.

4.4.2. Panel-Corrected Standard Error (PCSE) Approach

We estimated our model using the Panel-Corrected Standard Error (PCSE) technique in order to examine the connection between the independent and dependent variables. According to Doku and Kpekpena [91] and Nsanyan Sandow and Duodu [92], the PCSE technique effectively addresses the issues of autocorrelation, heteroskedasticity, and cross-sectional dependency. OLS yields skewed results when autocorrelation, heteroskedasticity, and cross-sectional dependence are present. The PCSE approach created by Beck and Katz [93] is considered a feasible alternative due to its ability to tackle the concerns identified by OLS. Based on the Monte Carlo simulation, Beck and Katz [93] suggested employing panel-corrected standard errors in PCSE instead of OLS standard errors. They also claimed that the PCSE estimator is very reliable when it comes to the efficiency gained by standard errors. The data were first altered to eradicate serial correlation in the PCSE method. Subsequently, the modified data underwent Ordinary Least Squares (OLS) analysis, where the standard errors were adjusted to account for cross-section dependence, heteroskedasticity, and autocorrelation. This process finally enhances the efficiency of estimation. A comparable function to that of PCSE is also fulfilled by the Feasible Generalized Least Squares (FGLS). But for FGLS to work, time (T) has to exceed cross-section (N). Given that N is larger than T in our panel dataset, PCSE is the most suitable method. Famanta and Randhawa [94]; Wang and Padhan [95]; and Magazzino [96] applied a similar PCSE methodology to check the environmental quality on different variables.

5. Results

Table 3. Descriptive Statistics.

Indicator	Mean	Std. Dev.	Min.	Max.	Data Source
Total Ecological Footprint (EF) (gha)	3.124	2.534	0.084	13.859	GFN (2022)
Inclusive Growth (IG) (0–100)	50,333.58	41,049.85	5154.55	179,303.3	WDI (2022)
Green Technological Innovation (GTI)	47,156.22	229,573.9	−7872	1,993,799	WDI (2022)
Total Fishery Production (TFP)	5,218,167	1.58 × 107	5141	8.80 × 107
Agriculture, Forestry, and Fishing (AFF)	8.98 × 1010	2.02 × 1011	7.94 × 107	1.22 × 1012	WGI (2022)
Renewable Energy Consumption (REC)	21.52985	22.48479	0.01	82.87	OECD (2022)
Aquaculture Production (AP)	3,456,502	1.25 × 107	0	7.48 × 107	WDI (2022)

describes the descriptive statistics of the variables used in this study. These variables comprise the ecological footprint (EF), inclusive growth (IG), green technological innovation (GTI), total fishery production (TFP), agriculture, forestry, and fishing (AFF), renewable energy consumption (REC), and aquaculture production (AP). It is shown from the results that, for all variables, the mean values are positive. The max. and min. values of the EF are 13.859 and 0.084. IG has the max. value (179,303.3) and min. value (5154.55). Similarly, GTI as well as TFP have max. values of 1,993,799 and 8.80 × 10⁷ and min. values of −7872 and 5141. The max. and min. values of AFF are 1.22 × 10¹² and 7.94 × 10⁷. REC’s max. value is 82.87, and the min. value is 0.01. However, AP has a max. value of 7.48 × 10⁷ and a min. value of 0.

Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7 show the trend of selected variables used in this study over time.

The results of CD, slope heterogeneity, and various diagnostic tests are shown in

Table 4. Cross-sectional dependence, slope heterogeneity, and diagnostic test results.

Econometric Issues	Diagnostic Tests	Test-Stat.	Prob.
CD	Breusch and Pagan LM	1983.00 ***	0.000
	Pesaran CD	4.312 ***	0.003
	Pesaran LM adj	0.6062	0.234
Slope heterogeneity	Swamy’s	48888.08 ***	0.000
	$\tilde{\Delta }$	3.532 ***	0.004
	${\tilde{\Delta }}_{\mathrm{adj}}$	4.665 ***	0.000
Autocorrelation	Wooldridge	32.223 ***	0.000
Heteroskedasticity	Modified Wald	7035.28 ***	0.000
	Breusch–Pagan/Cook–Weisberg	16.06 ***	0.000
Multicollinearity	Mean VIF	4.03	0.000

Note: *** level of significance: 1%; ** level of significance: 5%; and * level of significance: 10%.

. The Pesaran CD and Breusch and Pagan LM tests work well with smaller data, i.e., T > N, as in this panel study indicating CD’s existence in this panel of transition economies. On the other hand, the Pesaran LM adj test is applicable when N > T, which does not exist in this panel. Therefore, the Pesaran LM adj test describes no CD in this panel. Dos [97] said that interdependence arises when the economies of one or more nations are subject to the growth and development of other economies [98]. Slope heterogeneity was verified using the Swamy’s test, delta (Δ) test, and adjusted delta (Δ) test. The Wooldridge test indicates the existence of autocorrelation. The heteroskedasticity was confirmed using the Modified Wald test and Breusch–Pagan/Cook–Weisberg test. Nevertheless, because the VIF value is less than 5 (i.e., 4.03), multicollinearity does not exist.

The integration order for a few chosen variables (ecological footprints, inclusive growth, green technological innovation, total fishery output, agriculture, forestry and fishing, renewable energy consumption, and aquaculture production) is shown in

Table 5. CIPS unit root test.

Variables	At Level (Intercept and Trend)	At First Difference (Only with Intercept)
LnEF	−3.222 ***	−4.233 ***
LnIG	−4.336 ***	−3.667 ***
lnGTI	−0.889	−3.027 ***
lnTFP	−4.222 ***	−5.333 ***
LnAFF	−4.223 ***	−4.245 ***
LnREC	−5.888 ***	−6.55 ***
LnAP	3.567 ***	−4.234 ***

Note: *** level of significance: 1%; ** level of significance: 5%; and * level of significance: 10%.

using the CIPS unit root test. Two cases such as (a) at level with the intercept and trend (b) at first difference with the intercept alone were tested using this method. At a level and the first difference at a 1% significance level, all variables become stationary. However, only GTI is static at the first difference and non-stationary at a level. The stationarity existence implies the reliability of the regression coefficients.

A long-term relationship between the variables must be established to prevent misleading regression. Because of CD, the Westerlund cointegration test [80] was used. A long-run cointegration between the variables is confirmed in

Table 6. Westerlund cointegration test.

Panel	Variance Ratio
Westerlund (2005) Test	Statis.	Prob.
	−3.987 ***	0.0004

Note: *** level of significance: 1%.

. Due to statistically significant test results, the null hypothesis (no cointegration) was rejected.

Driscoll–Kraay standard error regression findings along with panel-corrected standard errors are shown in

Table 7. Regression with Driscoll–Kraay (D/K) standard errors (dependent variable: lnEF).

Variables	D/K Regression			PCSE
	Coff.	Std. Er.	Prob.	Coff.	Prob.
LnIG	−0.785 ***	0.0417	0.000	−0.571 **	0.000
lnGTI	−0.245 ***	0.023	0.000	−0.386 *	0.001
lnTFP	0.233 ***	0.0505	0.000	0.339 ***	0.000
LnAP	−0.046 *	0.0038	0.000	−0.596 ***	0.001
lnREC	−0.182 ***	0.045	0.000	−0.393 **	0.031
lnNRD	0.044 ***	0.183	0.000	0.356 ***	0.001
LnAFF	0.589 *	1.105	0.006	0.740 *	0.000
F-Stat.	14908.490 *** (0.000)			---
R2	0.846			---

Note: *** level of significance: 1%; ** level of significance: 5%; and * level of significance: 10%.

. Both types of results give almost similar results. The results show that inclusive growth (IG) negatively and significantly affects the ecological footprint (EF). A 1% rise in IG leads to a 0.785% and 0.571% fall in the EF at 1% and 5% levels of significance. Similarly, green technological innovation (GTI) also has a detrimental and considerable influence on the EF. A 1% rise in GTI leads to a 0.245% and 0.386% decrease in the EF at 1% and 10% significance levels. Conversely, total fishery production (TFP) significantly and positively impacts the EF. At the 1% significance level, a 1% increment in TFP causes a 0.233% and 0.339% increase in the EF. Both aquaculture production (AP) and renewable energy consumption (REC) harm and significantly affect the EF. A 1% increase in AP and REC puts 0.046% and 0.182% (10% and 1% significance level), 0.596%s and 0.393% (1% and 5% significance level) decreases in the EF. On the other hand, natural resource depletion (NRD) and agriculture, forestry, and fishing (AFF) positively and significantly influence EF. Through a 1% rise in NRD and AFF, the EF increases by 0.044% and 0.589%, and 0.356% and 0.740% at 1% and 10% levels of significance. The model accounts for 84.6% of the variation in the EF in these transition economies according to the R-square of 0.846.

Table 8. Results of sub-period D/K regression (robustness analysis).

Variables	D/K Regression
	Coff.	Std. Er.	Prob.
LnIG	−0.762 **	0.081	0.002
lnGTI	−0.446 ***	0.018	0.004
lnTFP	1.134 ***	0.135	0.000
LnAP	−0.044	0.201	0.522
lnREC	−0.383 ***	0.151	0.003
lnNRD	0.055 ***	0.005	0.003
LnAFF	0.423 ***	0.282	0.000
F-Stat.	1884.65.81 *** (0.000)
R2	0.973 *** (0.000)

Note: *** level of significance: 1%; ** level of significance: 5%; %; and * level of significance: 10%.

displays the robustness analysis (sub-period D/K regression) findings. The results show that IG negatively and statistically significantly affects the EF. At a 5% significance level, a 1% rise in IG leads to a 0.762% increase in the EF. Likewise, GTI has a deleterious and substantial impact on the EF. The EF decreases by 0.446% at a 1% level of significance for a 1% increase in GTI. Conversely, TFP has a substantial and favorable effect on the EF. With a significance level of 1%, it can be seen that a 1% increment in TFP lead to a 1.134% increase in the EF. Both AP and REC have a negative impact on the EF, whereas AP has an insignificant effect on the EF. At the 1% level of significance, a 1% boost in REC causes in a 0.383% drop in the EF. In contrast, it can be shown that NRD and AFF have a positive and statistically significant impact on EF with a significance level of 1%. By increasing NRD and AFF by 1%, the EF rises by 0.055% and 0.423%, respectively. The R-square value of 0.973 indicates that the model explains 97.3% of the variability in the EF in these transition economies.

6. Discussion

The results show that an increase in inclusive growth decreases ecological footprint. This is because, initially, the increase in inclusive growth causes an increase in environmental degradation but, in the long run, the increase in inclusive growth causes environmental improvement. Additionally, it encourages more environmentally friendly behaviors by guaranteeing a fair allocation of resources and financial prospects, which may result in lower consumption habits and a stronger emphasis on environmental preservation among all societal groups. The findings are consistent with [11,99]. Green technological innovation plays a considerable and favorable part in lessening the ecological footprint of transition countries. This view is supported by several factors. The industrial sector, which is primarily responsible for economic exports, is progressively shifting its technology from conventional (non-renewable) energy sources to more modern and clean (renewable) energy sources, hence reducing pollution levels overall. According to Usman and Hammar [52], the development of environmental technology contributes to a decrease in the utilization of fossil fuels, hence lowering energy consumption and promoting sustainable development. More technological advancements in the targeted nations seem to improve environmental results while reducing their ecological footprint. Numerous investigations including Zeraibi, Balsalobre-Lorente [100] and Usman, and Balsalobre-Lorente [101] confirmed these results. This study’s results align with the part of the body of research on the connection between the ecological footprint and agriculture, forestry, and fishing. Deforestation plans to make way for agriculture and forestry reduce biodiversity and exacerbate the negative impacts of both industries [102]. According to the current analysis, the ecological footprint is increased when agricultural land is used for farming. This increases the impact of forestry and agricultural production on the environment. The results show that total fishery production also increases the ecological footprint. The reason for this is that the act of catching fish, especially when performed extensively, can have a negative impact on marine ecosystems through practices like habitat destruction, bycatch, overfishing, and the use of fishing vessels that consume a lot of fuel that in turn increases environmental stress and the ecological footprint of the ecosystem [103,104]. According to research conducted by Muriithi [105], for example, the ecological footprint of Bangladesh is positively impacted by agriculture, forestry, and fishing combined with total fishery production. On the other hand, aquaculture production reduces the ecological footprint. By employing sustainable techniques including circular aquaculture systems, utilizing renewable energy, lowering N₂O emissions, preserving blue carbon, and utilizing efficient equipment, aquaculture can lessen its ecological footprint [106]. Yoshida and Lee [107] found that aquaculture production in South Korea hurts the ecological footprint.

Moreover, it is shown that there is a negative correlation between renewable energy consumption (REC) and the ecological footprint. It has been shown that, in transition nations, the REC coefficient has a negative significance for the EF in this respect. Increases in the use of renewable energy in transition countries show that efforts are being made to reduce environmental damage. Numerous scholars have contributed their insightful thoughts and perspectives on a variety of topics pertaining to environmental concerns and the use of renewable energy sources. Energy from renewable sources is expected to have a smaller environmental impact. These anticipated results are in line with the findings of Ozcan and Apergis [108], Usman and Radulescu [109], Saqib and Ozturk [110], and Usman and Balsalobre-Lorente [101], who confirmed that the use of renewable energy may achieve a long-term reduction in ecological footprint levels. The natural resource depletion puts a positive influence on ecological footprint. This conclusion is supported by the empirical research conducted by Alvarado and Tillaguango [111]; and Huang and Sadiq [112]. It could be due to the abundance of coal, forests, minerals, gas, oil, and other resources in the transition countries that have fueled industrialization processes, which have caused considerable environmental issues. The quest for economic advancement in these nations has led to a rise in the improper use of these natural resources, which has produced a significant amount of toxic waste. The natural resource depletion coefficient’s positive value recommends that, to have an economy that has an impact on the environment, nations with limited natural resources need to import fossil fuel energy such as petrol or gas. The results are related with the conclusions of [113].

7. Conclusions

The current study examines the relationships between the blue economy, environmental sustainability, natural resource depletion, renewable energy consumption, inclusive growth, and green technological innovation. This study uses panel data from 2000 to 2022 for 17 of the 24 transition nations. To check cross-sectional dependency in the countries, many cross-sectional dependence tests were used, including Breusch and Pagan LM, Pesaran CD, and Pesaran LM adj. In addition, the CIPS unit root test was used to check the stationarity of the variables. The Westerlund cointegration test results confirm that there is a strong correlation between the ecological footprint; inclusive growth; total fishery production; agriculture, forestry, and fishing; aquaculture production; green technological innovation; renewable energy consumption; and the depletion of natural resources. The long-term coefficient between the variables is then examined using the Driscoll–Kraay standard error and panel-corrected standard errors. The results show that aquaculture production, green technology innovation, inclusive growth, and the use of renewable energy eventually lower the ecological footprint. Additionally, in the long run, the production of fisheries as a whole, the depletion of natural resources, and agriculture, forestry, and fishing all improve the ecological footprint.

8. Policy Implication

This study delivers actionable recommendations for both policymakers and practitioners aiming to bolster environmental sustainability. It emphasizes the importance of improving environmental performance, channeling investments into renewable energy infrastructure, enacting supportive government policies and incentives, fostering green technological advancements, and adopting renewable energy sources. This research elaborates on the need for the promotion of an inclusive growth policy and technological innovation that could help in mitigating the ecological footprint in transition economies. The green practices and technologies that must be encouraged are investments in renewable energy and sustainable fishing, among others. In this view, such operations should embrace sustainable practices that allow the distribution of benefits at hand and the ecological impact of the business. This may include the adoption of clean technologies, responsible aquaculture, and the adoption of sustainable fisheries management.

Social Implications: The findings bring to light a social need that the blue economy needs to be socially sustainable, such that it benefits every other member in society, not only those in the top echelons but also those at the bottom, especially coastal communities. Social equity and the inclusion of all sections of society through blue economy initiatives are what can sustain, in the long run, the social well-being accrued through the blue economy.

Implication for further studies: This study further opens the ways for future studies of this nature to be carried out, which will help in exploring the intricate relationships of the blue economy with inclusive growth and environmental sustainability in different contexts. Moreover, it is due to further penetration into the dynamics of the blue economy with implications for society and the environment through longitudinal studies and comparative analyses between regions.

9. Research Limitations

9.1. Data Limitations

This study is secondary data-based, and the data are drawn from transition economies; therefore, it might not bring out some subtleties in the blue economy as they relate to the impacts on inclusive growth and environmental sustainability.

9.2. Methodological Limitation

The Driscoll–Kraay standard error regression approach is robust but, after all, can fail in accommodating all possible forms of cross-sectional dependence and heteroscedasticity contained in the panel data.

9.3. Generalizability

The study sample, composed of 17 transition economies, might impede the generalization of this study’s findings to other contexts or regions.

9.4. Temporal Scope

With the temporal value associated with the blue economy and its impacts even in the subsequent period, the frame within which the value is to be determined is accordingly set between 2000 and 2022.

First, in this paper, various sources of secondary data were used, and secondary data might contain measurement errors and consistency. Thus, there is a need to see that the similar findings prevail using primary data, which is something future research should use. Second, this study applied the analysis of panel data and cointegration tests, which can sometimes overlook the diversity and changing conditions unique to each country. Future investigations are anticipated to employ more sophisticated methodologies for examining causality and the dynamic interactions among variables, for instance, through the use of Panel Vector Autoregression (PVAR) or the Panel Vector Error Correction Model (PVECM). Additionally, this research considered merely six independent variables alongside a single dependent variable. Thus, subsequent studies are encouraged to encompass a more extensive selection of variables and factors.

10. Future Recommendations

A few suggestions have been made for further research that might build upon and enhance the analysis and conclusions of this investigation. Firstly, to improve the accuracy and reliability of the findings, future research may make use of more current and precise data sources for some of the factors, such as the utilization of renewable energy and the ecological effect. Secondly, other factors that might impact inclusive growth, such as innovation, human capital, and institutional quality, could be included in future research to enhance the model’s explanatory capacity and policy relevance. Thirdly, more advanced techniques, like the instrumental variable technique, panel vector autoregression, or panel data causality tests, can be used in future studies to tackle potential endogeneity, reverse causality, or omitted variable bias problems. These techniques could help to establish a stronger causative link among the independent and dependent variables.