The Impact of the Environment, Digital–Social Inclusion, and Institutions on Inclusive Growth: A Conceptual and Empirical Analysis

Abstract

Though the literature on inclusive growth is rich, further well-founded studies are required on the issue of sustainable inclusive growth. This paper seeks to summarize the role of environmental degradation in inclusive growth based on carbon dioxide emissions, and its interaction with factors such as social inclusion, digital inclusion, and institutions. One of the findings derived from the generalized method of moments (GMM) model is that sustainable inclusive growth can be achieved in all three income groups from the global data while focusing on institutional quality, digital inclusion, and social inclusion. Simultaneously, the harmful effects of carbon dioxide emissions can be circumscribed. The major recommendations of this study are that efforts to achieve sustainable inclusive growth should combine mutually reinforcing policies, namely: (i) promoting environmentally focused sustainable inclusive growth with socio-digital inclusivity; (ii) ensuring a strong institutional playing field for achieving inclusive growth; and (iii) strengthening macroeconomic policies, which means controlling inflation, and enhancing trade openness and literacy levels.

1. Introduction

Over the next few years, it is assumed that the increasingly globalized world will experience tremendous economic growth, despite the current serious challenges that it faces. Making progress means overcoming the interconnected issues of environmental degradation, socio-digital inclusion, and inclusive growth. First, the world needs to meet the challenges associated with growing concerns about the state of the environment [1,2]. It is projected that the world’s increasingly unsustainable population will increase from 7 billion in 2012 to 9.3 billion by 2050 [3]. Almost 50% of the population is growing in the Sub-Saharan Africa region alone [4]. With the ever-increasing number of people, countries will fight for food, water, jobs, natural resources, and shelter at the cost of damaging the environment [5]. The demand for more resource-intensive foods such as meats and vegetable oils will grow, and it requires intensive measures to protect the environment. Unchecked population growth will also increase industrial activities, carbon dioxide, and other harmful gas emissions, leading to depletion of the ozone layer and, subsequently, environmental degradation [6,7,8]. Such harmful gasses in environments, coupled with growing industrialization, urbanization, and populations, has serious repercussions for all humans, flora, and fauna [9,10]. Currently, carbon dioxide emissions are rising at a fast rate globally, posing a major threat to all forms of life (see Figure 1).

Secondly, it is reported that roughly 870 million of the world’s poorest people remain undernourished today, and there is no doubt that the situation will only worsen in the coming years [11]. Many millions of people are already living their lives on the margins of poverty, as evidenced by food riots in 2008 which broke out in nearly twenty countries [12]. Poverty is so severe that to feed all people by the year 2050, worldwide food availability must increase by 64% [13]. With the growing population and changing environmental dynamics, agriculture will be seriously hit by these environmental crises, affecting food availability, and creating a spike in food prices. Agriculture accounts for nearly 3% of global GDP and employs almost two billion people around the world. Agriculture is a major source of greenhouse gas emissions and one of the largest consumers of water. With the growing population, if employment opportunities are not generated at a matching pace, the goal of achieving inclusive growth will never be realized. The result will be a sharp increase in poverty and widening income inequalities.

Third, the dramatic consequences of environmental degradation are leading to uncontrolled emissions, which will produce tremendous increases in the rate of global warming, particularly in the next few years [14,15,16,17,18]. Economic growth accompanied by huge environmental damage will be inevitable, and such degradation will bring about serious health concerns, limiting any meaningful notion of inclusive economic growth. Environmental degradation will reduce people’s lifespans and their economic progress will not be possible. The first question we ask in this study is: do carbon dioxide emissions contribute to discouraging inclusive growth, particularly in different income groups? The second question that we aim to investigate is: how do carbon dioxide emissions affect growth in social and digital inclusion within different regional groups? To answer these queries, we used a world panel dataset comprising countries from different continents, with the results dependent on how much data could be gathered for the purposes of this research.

2. Literature Review

In contemplating the future of inclusive growth in different global regions, economics and environmental science dictate the serious consequences that can undermine a country’s ability to achieve inclusive growth. It is usually perceived that economic growth and its related activities are largely affected by processes in agricultural, and in the industrial and energy sectors [19,20]. Among these three, the latter is perhaps the more significant and is discussed in this study. It is impossible to achieve sustainable inclusive growth without focusing on the world’s climate problems that are the outcome of decades, if not centuries, of industrialization, especially in the decades since World War II [21,22]. While referring to climatic changes, we generalize the average characteristics of the atmosphere in different regions of the world, which may vary according to diurnal episodes. More precise representation of the climatic changes that are occurring serves as an accumulative indicator of environmental degradation. However, data on all indicators on a global level are not readily available on any genuine data sites. Considering data availability, the chief variable for which data are widely available is carbon dioxide emissions [14,21,23].

Carbon dioxide emissions have the most significant influence on the world’s climate. Many climatologists and economists feel that it is prudent to consider this fact on a global level, given what has happened during the last century. Carbon dioxide emissions are largely the result of the combustion of fossil fuels [24,25,26,27]. Once these emissions become part of the atmosphere around us, the residence time of these emissions in air appears to be long enough when compared to industrial carbon dioxide existence in air [27,28,29]. It is predicted that ever-increasing carbon dioxide emissions will elevate the global temperature by almost three degrees Celsius in the decades to come [30,31]. However, this predicted change largely depends on how severe conditions become within the next few years if something is not done about them.

Inclusive growth seeks to generate productive employment opportunities and the curtail poverty and income equalities, but it is only of interest to people if they are healthy and have many years to live [32,33,34,35,36]. It is important to state here that although many studies have been conducted on this issue, the problem of controlling carbon dioxide emissions is still something that environmental policymakers cannot master consistently [37]. The problem is the implementation of environmental control policies on a more decentralized level, which requires a strong institutional structure and enforcement procedures, which developed countries have [38,39,40]. Policies will generate results if they are acceptable to the wider society, if law and order, accountability, and crackdown on corruption are maintained, and if regulations are properly implemented [41,42,43]. Unfortunately, many developing countries are experiencing environmental degradation, and there are no proper enforcement agencies [44]. Weak institutional structure is at the mercy of nepotism, corruption, political favors, poor quality regulations, etc. [45,46,47,48,49,50]. Moreover, external conditions cannot be controlled by domestic market or political mechanisms, despite the need to produce policies that control carbon dioxide emissions.

The world needs to also reduce agriculture’s impact on the environment and natural resources, and not just focus entirely on industrial waste [51,52]. Agriculture is a major contributor to greenhouse gas emissions, the largest industry consumer of freshwater, and also the largest cause of damage inflicted upon natural ecosystems [53,54,55]. In the future, agriculture will need to embrace new technologies that ensure adequate food production and are environmentally friendly [56,57,58,59]. Agriculture can provide numerous benefits beyond food production and jobs, including building materials for soil fertility, and more [60].

The need for agriculture to support to food security and industry, and to ensure a healthy environment, poses one of the paramount challenges for the next several decades: how can the world adequately and fairly feed a growing population in a way that alleviates poverty and makes economic progress possible while, at the same time, reducing pressure on natural resources? This research paper provides several perspectives on answering this question, which has been overlooked in the above-cited literature. Through a number of core propositions, it makes the case that integrating sustainability considerations into a post-COVID-19 era on environmental changes will be critical to achieving inclusive growth. The paper is based on empirical findings, and it will conclude by recommending several policies targets/guidelines―along with their associated indicators and means of implementation―that would incorporate some important sustainability considerations into achieving inclusive growth.

Another issue that is closely related to sustainable inclusive growth is the low level of digital and social inclusion, particularly in the developing world, which keeps people isolated from understanding environmental issues [61,62,63,64]. Building social cohesion and inclusion, and addressing social exclusion and inequality, is important for several different reasons: First, it makes sure that the most vulnerable and marginalized are supported, which is important in and of itself and also aids in conflict mitigation. Second, groups of people who feel even more excluded and marginalized often feel less responsible for the environment around them and care less about the needs of people around them. Social and digital inclusion keeps people connected and informed about recent global events, and thus, they obtain some understanding of what inclusive growth entails.

Institutions are the key drivers of not just economic growth, but also sustainable inclusive growth [41,65,66,67,68]. Strong institutional structures, such as those in developed countries, create stability, and secure rights for individuals to become engaged in economic transactions, investment, education, contracting, and legal functions [69]. Institutions lead to technological change, innovations, and productivity, as well as various aspects of shared prosperity. Inclusive economic institutions are both incentive-related and opportunity related. So, secure and stable environments are good incentives for people to not just undertake investments or better themselves, but also to find new ways of approaching new problems or existing problems such as environmental damage. At the same time, economic factors cannot be separated from political and social ones, or institutions. A political system where power is vested in the hands of a few will not be beneficial for the economy or target inclusive growth variables such as productive employment opportunities, poverty reduction, the elimination of inequalities, or the simple achievement of higher economic growth. The reason is that economic institutions are shaped by political institutions and the distribution of political power in society. This relationship is so organic that it is almost impossible for economic institutions to remain inclusive when political institutions and the political structures are very much against inclusivity or sustainable inclusive growth that takes the environment into account. Even though institutions and sustainable inclusivity is so tightly knitted to each other, hardly any significant studies have explored the impacts of environmental degradation on inclusive growth, which takes care of institutional structures in a world-level-data setting.

Hardly any significant studies have explored the above contexts, particularly from the perspective of environmental degradation and inclusive growth with global data divided into three income groups. Unfortunately, people in under-developed areas barely recognize the importance of the environment, and future generations will suffer consequently [69,70,71]. Being connected to the internet (digitally included) can raise awareness among people in under-developed regions about environmental degradation and the need for sustainable inclusive growth [72,73,74]. In this seminal work, we stress that it is not only inclusive growth that is required, but “sustainable” inclusive growth which promises a healthier environment for current and future generations. This paper attempts to fill this research gap by providing a thorough analysis of the role of environmental degradation, along with that of institutions and socio-digital inclusion, in achieving inclusive growth in three income groups.

3. Conceptual Framework

Inclusive growth is linked to several macroeconomic conceptual frameworks. In this section, we develop a relatively new macro-economic concept which encapsulates the inherent meaning of inclusivity while also considering other concepts, which are stated below. All the concepts presented below will then, together, form the main conceptual framework of sustainable inclusive growth. The first concept of inclusive growth is centered on the “well-being concept”. Well-being refers to improving the quality of not only human-to-human, but also human-to-ecology interaction [75]. This can be assisted by new technologies that make sustainable economic development possible. Well-being economy focuses on horizontal growth, unlike the ‘trickledown’ growth assumption, which follows a vertical growth structure focused on the separation of production and consumption, and only helps people already on higher incomes and their well-being [76].

On the other hand, consumption-based development models depend on consumers increasing their achievement of well-being by purchasing goods and services, and this follows vertical growth patterns [77]. Rising consumption would enable people to make more objective decisions linked with their self-esteem, self-value, and personal motives. The fundamental rationale of the “consumption concept” is to achieve growth through consumption and the standardization of wants and needs [78]. It encourages society to expand on individual and collective interests while also caring for the environment, which is vital for our existence.

The second concept that relates to inclusive growth is the “sharing economy”. Shared growth is one of the best definitions of inclusive growth [79]. In this concept, the focus is on collectivistic consciousness and growth through the community making progress [80]. We are social animals who thrive when our relationships with others are equal. Owing to our relationships, we also build a relationship of care for the ecosystem. Low-impact processes of the generation of goods and services focus on drastically reducing waste and strengthening interpersonal relationships so that production is inclusive. Technological revolutions play a key role in the sharing economy by building collaborations among people [81]. Social networks are the key to the sharing economy and explain how, to some extent, the world is now a global village [82]. Increasing the productive participation of all members of a society creates a sense of collective ownership and responsibility towards each other in terms of achieving economic growth, without compromising on environment standards.

The third concept that relates to inclusive growth is the “saving economy” [83,84]. This idea is very different to the consumption model, with a focus on saving, which is an important constituent of inclusive growth. Saving the environment is an important aspect of this concept. Future generations will be able to live in a better world when the economy concentrates on saving what is important, which creates links with sustainable inclusive growth [85].

Another concept that is part of inclusive growth is the “access economy”. This concept assumes there is a fully horizontal process of sharing goods and services made possible by direct trade [86]. There are direct links among consumers, but this type of economy is dictated by fluctuations in the prices of goods/services, which could lead to inflation. Inflation hurts people’s purchasing power in the long run. People feel poorer and the struggle to acquire even the basic needs compromises the core idea of inclusive growth. The last concept related to inclusive growth is the notion of the “economics of sharing”. Sharing involves caring for others in terms of growing inclusively [87]. This concept is based on shared public ethics in terms of reducing unemployment and poverty [88]. Unlike the concept of social inclusion, this theory values the independent contribution of human beings in receiving and sharing resources from nature. A sense of responsibility is important to preserve nature and the environment. Over-consumption and over-production result in waste and carbon dioxide emissions, which harm the ecosystem [89]. If care is not taken here, not only is there huge economic loss, but inclusive growth is also not possible.

We developed conceptual links among socio-digital inclusion, environment (measured via CO₂ emissions), and inclusive growth, as depicted in Figure 2’s conceptual framework. It is based on social conscious capital, which is the product of social inclusion. The capital required for inclusive growth is not merely goods or services, but the value of resources, which depends on our interrelationships and those with the natural world. The concepts emphasize the humanitarian, community perspective, modern environmentalism, and individualism, to ensure sustainable inclusive growth. A shift in thinking paradigms is required to realize this. We elaborate the need for strengthening social and digital ties to raise people’s awareness about environmental needs. In a nutshell, inclusive economic growth focuses on a system of social relationships and the ecosystem, which gives the notion of collaborative and sustainable inclusive growth real meaning.

4. Methodology and Model Specifications

Dynamic panel macroeconomic models explore the relationship between variables while taking the lagged effect of dependent variables into account. This inclusion also takes care of any misspecification biases in the model. We have employed the two-step system GMM model to conduct our empirical analysis. The general functional form of our model to be estimated is:

$$IG=f\left(Instit,SI,DI,{CO}_2, Trade, INV, EDU, INF\right)$$

(1)

In Equation (1) above, IG, Instit, SI, DI, CO₂, Trade, INV, EDU, and INF represent, respectively, inclusive growth, institutions, social inclusion, digital inclusion, carbon dioxide emissions, trade openness, investment, education, and inflation. Meanwhile the econometric equations may be written as:

• Model 1

$$\begin{array}{c}IG{ }_{it}={\gamma }_0+{\gamma }_1 Instit{ }_{it}+{\gamma }_2 SI{ }_{it}+{\gamma }_3 D{I}_{ it}+{\gamma }_4 {CO}_2{ }_{it}+{\gamma }_{5 }Trade{ }_{it}\hfill \\ \hfill \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\gamma }_6 IN{V}_{ it}+{\gamma }_7 Edu{ }_{it}+{\gamma }_8 Inf{ }_{it}+{\gamma }_9 Lag.IG{ }_{it}+error{ }_{it}\hfill \end{array}$$

(2)

• Model 2

$$\begin{array}{c}IG{ }_{it}={\gamma }_0+{\gamma }_1 Instit{ }_{it}+{\gamma }_2 SI\ast {CO}_2{ }_{it}+{\gamma }_3 S{I}_{ it}+{\gamma }_4 {CO}_2{ }_{it}+{\gamma }_{5 }Trade{ }_{it}\hfill \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\gamma }_6 IN{V}_{ it}+{\gamma }_7 Edu{ }_{it}+{\gamma }_8 Inf{ }_{it}\hfill \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\gamma }_9 Lag.IG{ }_{it}+{\gamma }_{10} DI{ }_{it}+error{ }_{it}\hfill \end{array}$$

(3)

• Model 3

$$\begin{array}{c}IG{ }_{it}={\gamma }_0+{\gamma }_1 Instit{ }_{it}+{\gamma }_2 DI\ast {{CO}_2}_{it}+{\gamma }_3 D{I}_{ it}+{\gamma }_4 {{CO}_2}_{it}+{\gamma }_{5 }Trade{ }_{it}\hfill \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\gamma }_6 IN{V}_{ it}+{\gamma }_7 Edu{ }_{it}+{\gamma }_8 Inf{ }_{it}\hfill \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\gamma }_9 Lag.IG{ }_{it}+{\gamma }_{10} SI{ }_{it}+error{ }_{it}\hfill \end{array}$$

(4)

• Model 4

$$\begin{array}{c}IG{ }_{it}={\gamma }_0+{\gamma }_1 Instit\ast {CO}_2{ }_{it}+{\gamma }_2 SI{ }_{it}+{\gamma }_3 D{I}_{ it}+{\gamma }_4 {CO}_2{ }_{it}+{\gamma }_{5 }Trade{ }_{it}\hfill \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\gamma }_6 IN{V}_{ it}+{\gamma }_7 Edu{ }_{it}+{\gamma }_8 Inf{ }_{it}+{\gamma }_9 Lag.IG{ }_{it}\hfill \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} ++{\gamma }_{10} Instit{ }_{it}+error{ }_{it}\hfill \end{array}$$

(5)

Equation (2) considers the impact of social inclusion, digital inclusion, institutions, and carbon dioxide emissions, along with other control variables, on inclusive growth. Conversely, Equations (3)–(5) look for interactions between social inclusion, digital inclusion, and institutions with carbon dioxide, respectively. Models 1, 2, and 3 are run on three different samples of high-income, low-income and middle-income countries to generate a comparative analysis. This is given in detail in

Table A1. World ranking on the basis of income groups.

Rank	Country	Economic Status	Rank	Country	Economic Status
1	Finland	High Income	43	Ghana	Lower-Middle Income
2	Norway	High Income	44	Brazil	Upper-Middle Income
3	Sweden	High Income	45	Jordan	Lower-Middle Income
4	Luxem.	High Income	46	Tunisia	Lower-Middle Income
5	Netherlands	High Income	47	Indonesia	Lower-Middle Income
6	Australia	High Income	48	Ukraine	Lower-Middle Income
7	Iceland	High Income	49	Honduras	Lower-Middle Income
8	UK	High Income	50	Colombia	Upper-Middle Income
9	Denmark	High Income	51	Tanzania	Low Income
10	Austria	High Income	52	Uzbek.	Lower-Middle Income
11	Ireland	High Income	53	Malawi	Low Income
12	Belgium	High Income	54	Kyrg.	Lower-Middle Income
13	US	High Income	55	Kazakhstan	Upper-Middle Income
14	Japan	High Income	56	Algeria	Upper-Middle Income
15	Cyprus	High Income	57	Lebanon	Upper-Middle Income
16	Poland	High Income	58	Sri Lanka	Lower-Middle Income
17	Slovakia	High Income	59	Nepal	Low Income
18	Portugal	High Income	60	Bangladesh	Lower-Middle Income
19	Chile	High Income	61	Bolivia	Lower-Middle Income
20	Hungary	High Income	62	Vietnam	Lower-Middle Income
21	Czech Republic	High Income	63	Moldova	Lower-Middle Income
22	Spain	High Income	64	Thailand	Upper-Middle Income
23	Estonia	High Income	65	Iran	Upper-Middle Income
24	Italy	High Income	66	Turkey	Upper-Middle Income
25	Slovenia	High Income	67	Armenia	Lower-Middle Income
26	Croatia	Upper-Middle Income	68	Russian Federation	Upper-Middle Income
27	Korea	High Income	69	China	Upper-Middle Income
28	Lithuania	High Income	70	Pakistan	Lower-Middle Income
29	Panama	Upper-Middle Income	71	Uganda	Low Income
30	Uruguay	High Income	72	Ecuador	Upper-Middle Income
31	Latvia	High Income	73	Madagas.	Low Income
32	Rwanda	Low Income	74	Georgia	Lower-Middle Income
33	Bulgaria	Upper-Middle Income	75	Azerbaijan	Upper-Middle Income
34	Malaysia	Upper-Middle Income	76	Yemen, Rep.	Lower-Middle Income
35	India	Lower-Middle Income	77	Belarus	Upper-Middle Income
36	South Africa	Upper-Middle Income	78	Nigeria	Lower-Middle Income
37	Romania	Upper-Middle Income	79	Paraguay	Upper-Middle Income
38	Philippi.	Lower-Middle Income	80	Egypt	Lower-Middle Income
39	Mexico	Upper-Middle Income	81	Iraq	Upper-Middle Income
40	Argentina	Upper-Middle Income	82	Togo	Low Income
41	Morocco	Lower-Middle Income	83	Zimbabwe	Low Income
42	Peru	Upper-Middle Income

Note: green color is for high income (HYC), yellow shows middle income (MYC), and red shows low income (LYC).

: distribution and division, of Appendix A for the sample countries.

4.1. The Generalized Method of Moments (GMM)

Endogeneity plays a significant role in producing bias in the estimators of conventional econometric tools. The simultaneity problem needs to be solved to obtain usable results. Ref. [90] formalized a generalized method of moments (GMM) strategy that does not require all the information on how the data are distributed. The GMM method is commonly used for treating endogeneity and heteroscedasticity. The dynamic panel estimators and the difference and system GMM [91,92] estimators are formalized especially for small-sample data where the number of cross-sections is greater than the period.

Suppose we have an econometric model equation:

$$Y=X^{\prime }\beta +ϵ$$

(6)

where the error term is independent of instrument variables $E\left(I\right)=0$. β is the coefficient vector. Y is the dependent variable and X denotes the column vector of k independent variables, $X={(x_1,x_2, \dots x_k)}^{\prime }$. I stand for the column vector of j instrument variables, $I={(i_1,i_2, \dots i_j)}^{\prime }$. X and I can share their elements (variables) because the moments of the regressors can be used as instruments, and j ≥ k. X, Y, and I are the matrices and x, y, and i are the variables. E = Y − Xβ and the estimated residuals are $\widehat{E}={(\widehat{{ϵ}_1},\widehat{{ϵ}_2},\dots ,\widehat{{ϵ}_N})}^{\prime }$, which can be written as $\widehat{E}={(\widehat{{ϵ}_1},\widehat{{ϵ}_2},\dots ,\widehat{{ϵ}_N})}^{\prime }$= Y − X$\hat{\beta }$. The necessary condition for the instruments being valid is orthogonality of the instrument to the residuals, $E\left(ϵ,I\right)=0$. Theoretically and empirically, $E_N\left(I\epsilon \right)=\left(\frac{1}{N}\right)I^{\prime }\widehat{E}$.

In the generalized method of moments, the magnitude can be found through a generalized metric consisting of a positive semi-definite quadratic function. Suppose we have Q, which is the matrix of that quadratic function. After that, the equation is written as follows:

$$\Vert E_N\left(Iϵ\right)\Vert Q=\Vert 1/N{I}^{\prime }\widehat{E}\Vert Q=N{(\frac{1}{N}{I}^{\prime }\widehat{E})}^{\prime }Q\left(\frac{1}{N}{I}^{\prime }\widehat{E}\right)=\frac{1}{N}\widehat{E^{\prime }}IQ{I}^{\prime }\widehat{E}.$$

(7)

To obtain the desired vector of coefficients ${\beta }_Q$, we need to minimize ${\beta }_Q=argmi{n}_{\widehat{\beta }}\Vert I^{\prime }\widehat{E}\Vert Q$, and ${\beta }_Q$ can be derived through $d/d(\widehat{\beta })\Vert I^{\prime }\widehat{E}\Vert Q=0$. By following the chain rule of derivatives, this equation can be expanded as set out below:

$$0=\frac{d}{d(\widehat{\beta })\Vert I^{\prime }\widehat{E}\Vert Q}=\frac{d}{d(\widehat{E})}\Vert I^{\prime }\widehat{E}\Vert Q\frac{d\widehat{E}}{d(\widehat{\beta })}$$

(8)

$$0=\frac{\frac{d}{d(\widehat{E})}\left\{\frac{1}{N}\widehat{E^{\prime }}(IQ{I}^{\prime })\widehat{E}\right\}d(Y-X\widehat{\beta })}{d\widehat{\beta }}=\frac{2}{N}\widehat{E}IQ{I}^{\prime }\left(-X\right)$$

(9)

After dropping -2/N and taking the transpose, Equation (4) becomes:

$$0=\widehat{E^{\prime }}IQ{I}^{\prime }X={(Y-\widehat{{\beta }_Q})}^{\prime }IQ{I}^{\prime }X=Y^{\prime }IQ{I}^{\prime }X-{\widehat{\beta }}_Q^{\prime }{X}^{\prime }IQ{I}^{\prime }X$$

(10)

$$X^{\prime }IQ{I}^{\prime }X\widehat{{\beta }_Q}=X^{\prime }IQ{I}^{\prime }Y$$

(11)

Equation (6) can be written as:

$$\widehat{{\beta }_Q}=\frac{X^{\prime }IQ{I}^{\prime }Y}{X^{\prime }IQ{I}^{\prime }X}={(X^{\prime }IQ{I}^{\prime }X)}^{-1}{X}^{\prime }IQ{I}^{\prime }Y$$

(12)

Good instruments are not available outside the model, so these estimates are obtained using internal instruments. Consequently, the general model of the dynamic panel models is:

$$y_{it}=\gamma y_{i,t-1}+x_{it}^{\prime }\beta +{\epsilon }_{it}$$

(13)

where ${\epsilon }_{it}={\vartheta }_i+{\omega }_{it}$, and also, $E\left({\vartheta }_i\right)=E\left({\omega }_{it}\right)=E\left({\omega }_{it},{\vartheta }_i\right)=0$. The error term contains two orthogonal parts. The first is the fixed part ${\vartheta }_i$ and the second part ${\omega }_{it}$ is idiosyncratic shocks. The equation can be written as:

$$\mathrm{\Delta }{y}_{it}=\left(\gamma -1\right)y_{i,t-1}+x_{it}^{\prime }\beta +{\epsilon }_{it}$$

(14)

4.2. Data and Sources

Based on the above equations, we carefully examine the effect of the determinants of inclusive growth by using the panel data for eighty-three countries, comprising the years 2010 to 2018. It is important to state here that carbon dioxide emission data are not available from the WDI after 2018. Since N = 83 and T = 8, and therefore, N > T, we can apply GMM. The countries are organized into high-income, middle-income, and low-income groups based on the World Bank’s classifications. The min–max normalized indexing technique constructs the four indices of social inclusion, institutions, digital inclusion, and inclusive growth. Social inclusion (SI), inclusive growth (IG), digital inclusion (DI), and the institutional index (Instit) are adopted from the recent works of Aslam et al. (2021) and Aslam and Shabbir (2019). The social inclusion index comprises 11 variables: “avoidance of homicides, legal rights, financial inclusion, mortality rate, expenditures on health and education, political rights, freedom, GDP growth, school enrollments, life expectancy at birth and in vulnerable employment”. For details see [48]. The inclusive growth index is a composite of 4 variables including: “GDP growth, no poverty, income equalities and employment to population ratio”. The digital inclusion index includes 3 variables, i.e., “the number of broadband connections, internet users and the number of mobile users”. Conversely, the institutional quality index is a composite of six institutional quality measures: “control of corruption, government effectiveness, rule of law and order, regulatory quality, voice and accountability, political stability and absence of violence”. These are adopted from: [45,46,47,48,49,50]. Moreover, the data on controlling variables (inflation, trade openness, and investment) and the variables used in building up the indices above (except the institutions) are acquired from the World Development Indicators (WDI), published by the World Bank. The data on institutions are sourced from the International Country Risk Guide (ICRG) and World Governance Indicators (WGI).

5. Results and Discussion

Table 1. Results for the impact of CO₂ on inclusive growth without interaction terms.

	High Income	Middle Income	Low Income
Inclusive Growth_1	0.5943 ***	0.4620 **	0.3084 ***
	(0.1331)	(0.1638)	(0.0582)
	[4.4637]	[2.8214]	[5.2973]
Institutions	3.4716 ***	3.3478 **	0.3284 ***
	(1.0232)	(1.6617)	(0.6561)
	[3.3929]	[2.0146]	[0.5006]
Social Inclusion	0.4557 *	0.3600 **	0.2370 *
	(0.2537)	(0.1324)	(0.1289)
	[1.7963]	[2.7189]	[1.8381]
Digital inclusion	0.6380 ***	0.1618 *	0.0962
	(0.1075)	(0.0956)	(0.1144)
	[5.9340]	[1.6920]	[0.8591]
CO2	0.1476	0.4309 ***	0.1067 ***
	(0.1126)	(0.0641)	(0.0207)
	[1.3106]	[6.7279]	[5.1447]
Trade Openness	0.2467 ***	0.2384 ***	0.1519 *
	(0.0512)	(0.0504)	(0.0828)
	[4.8172]	[4.7287]	[1.8336]
Investment	0.6867 ***	0.4568 ***	0.3048
	(0.1160)	(0.1379)	(0.6348)
	[5.9209]	[3.3135]	[0.4801]
Education	0.1984 **	0.0789 ***	0.0402 **
	(0.0999)	(0.0183)	(0.0178)
	[1.9858]	[4.3121]	[2.2578]
Inflation	−0.0181	−0.0689 *	−0.0632 ***
	(1.1173)	(0.0396)	(0.0156)
	[−0.0162]	[−1.7378]	[−4.0533]
Constant	31.356 ***	29.217 ***	27.137 **
	(6.2023)	(8.0475)	(11.018)
	[5.0556]	[3.6306]	[2.4628]
Diagnostics
Wald test (p value)	3613.4 *** (0.0000)	1942.8 *** (0.0000)	4691.3 *** (0.0000)
Arellano–Bond test AR(2) (p value)	1.3714 (0.9835)	0.8379 (0.6237)	2.7368 (0.3987)
Hansen Test for IV (p value)	2.3484 (0.2321)	1.7824 (0.1328)	2.2167 (0.9748)

Source:Estimations by the authors. The symbols *, **, and *** represent significance at the 10%, 5%, and 1% levels, respectively. The values in parentheses are standard errors and the values in square brackets are the t-values. Note: IV in Hansen test is Instrument Validity.

summarizes the results of the determinants of inclusive growth, considering the impact of carbon dioxide emissions. In the high-income countries, seven variables—institutions, social inclusion, digital inclusion, trade openness, investment, and education—significantly contribute to inclusive growth. The results agree with those of another research [46,48]. However, two variables, including carbon dioxide emissions and inflation, have an insignificant impact on inclusive growth in high-income countries. Carbon control environmental policies are very important in high-income and higher middle-income countries because they emit approximately 86% of global carbon dioxide emissions [93]. In contrast, low-income countries produce 14% of total carbon dioxide emissions. Interestingly, although the population is tremendously high in low-income countries, increase in population is not affecting much of the emissions. It is interesting to note that the poorest countries are a source of just 0.5% of carbon dioxide emissions [94]. These figures provide a strong indication of the relative sensitivity of global carbon dioxide emissions to various income countries versus population and poverty. Both poverty and population are elements of the inclusive growth index. Inflation, on the other hand, has an insignificant impact on inclusive growth because inflation is controlled in the developed world, and strong macroeconomic policies and stability of the exchange rate further help in this process.

Referring to the middle-income countries, all the variables significantly contribute to inclusive growth (Table 1). Carbon dioxide emissions have an important impact about analysis, though the outcome is positive. It is important to note that increases in emissions are the result of technological/industrial processes and urbanization. Hence, the impact of carbon dioxide emissions is positive. However, policies must be designed to curtail carbon dioxide emissions and not compromise the pace of innovation-led inclusive growth [28,72]. In the low-income countries, it is apparent that eight determinants significantly contribute to inclusive growth. The significance of inflation predicts the weak institutional structures which are supposed to control inflation and inclusiveness [95].

Table 2. Results for the impact of CO₂ on inclusive growth with social inclusion * CO₂.

	High Income	Middle Income	Low Income
Inclusive Growth_1	0.5919 ***	0.4984 **	0.4444 *
	(0.0769)	(0.2510)	(0.2543)
	[7.6976]	[1.9862]	[1.7477]
Institutions	4.4885 **	3.6975 *	0.0733
	(2.1878)	(1.9989)	(0.2462)
	[2.0516]	[1.8489]	[0.2976]
Social Inclusion	0.3278 *	0.3420 ***	0.3073 *
	(0.1771)	(0.0567)	(0.1710)
	[1.8511]	[6.0320]	[1.7969]
Social Inclusion * CO2	0.9685 *	0.4382 ***	0.3772 *
	(0.5732)	(0.1107)	(0.1044)
	[1.6898]	[3.9583]	[3.6141]
Digital inclusion	0.2753 **	0.2385 **	0.1545 *
	(0.1039)	(0.1197)	(0.0818)
	[2.6508]	[1.9928]	[1.8891]
CO2	0.0269	0.6188 ***	0.8955 ***
	(0.1037)	(0.1167)	(0.2338)
	[0.2594]	[5.3024]	[3.8307]
Trade Openness	0.6098 **	0.5131 ***	0.4608 *
	(0.2525)	(0.0996)	(0.2689)
	[2.4153]	[5.1526]	[1.7133]
Investment	0.2466 ***	0.3099 **	0.1282 *
	(0.0674)	(0.1062)	(0.0721)
	[3.6610]	[2.9176]	[1.7777]
Education	0.6431	0.5404 **	0.4912 **
	(0.8699)	(0.2428)	(0.6604)
	[0.7393]	[2.2255]	[2.2578]
Inflation	−0.1198	−0.3805 *	−0.0217 **
	(0.5519)	(0.2131)	(0.0093)
	[−0.2171]	[−1.7857]	[−2.3396]
Constant	30.831 ***	26.741 ***	24.076 ***
	(4.2150)	(2.9502)	(2.9125)
	[7.3145]	[9.0643]	[8.2665]
Diagnostics
Wald test (p value)	4184.1 *** (0.0000)	2294.7 *** (0.0000)	3992.6 *** (0.0000)
Arellano–Bond test AR(2) (p value)	2.2674 (0.8450)	1.5321 (0.3849)	1.4562 (0.1342)
Hansen Test for IV (p value)	1.4824 (0.8549)	2.4695 (0.8402)	1.4592 (0.3659)

Notes: The symbols *, **, and *** represent significance at the 10%, 5%, and 1% levels, respectively. The values in parentheses are standard errors and the values in square brackets are the t-values. Note: IV in Hansen test is Instrument Validity.

summarizes the impacts of various determinants of inclusive growth, which takes care of the interaction term of social inclusion and carbon dioxide emissions. The lagged term of inclusive growth is significant in all three income groups. Institutions are only significant in high-income and middle-income countries, but insignificant in low-income countries. The reason could be that institutions have a stronger impact in these income groups, while the quality of them is better in high- and middle-income nations [41,46,65].

Social inclusion is significant in all three income groups, but is highly significant in middle-income countries. Social inclusion is an imperative variable for inclusive growth [46,49]. The interaction of social inclusion and carbon dioxide emissions is significant in all three cases. This predicts the validity and rationale of the interaction term. The logical interpretation of such results could be that social inclusion increases interaction among people and their awareness of ongoing changes. Particularly in this case, social inclusion helps to control the rate of environmental degradation while targeting inclusive growth. However, carbon dioxide emissions affect inclusive growth in all cases except the high-income group. Industrial development leads to inclusive growth, which comes with the cost of increased carbon dioxide emissions. Thus, it shares a positive relationship of carbon dioxide with inclusive growth, although carbon dioxide is undesirable when we target sustainable inclusive growth.

An important point that emerges from such results is that carbon dioxide may insignificantly affect inclusive growth in high-income countries but the interaction between social inclusion and carbon dioxide emissions can greatly achieve inclusive growth. Investment and trade openness are important in all three cases. Meanwhile, the impact of education is only significant in the middle-income and high-income countries. In the latter, people are already highly literate, and hence, its impact may be insignificant. Lastly, inflation negatively affects inclusive growth in all three cases, but it is significant in middle-income and high-income countries only. These results are indicative of weak microeconomic policies and higher economic instability in these regions [96,97].

Table 3. Results for the impact of CO₂ on inclusive growth with digital inclusion * CO₂.

	High Income	Middle Income	Low Income
Inclusive Growth_1	0.5768 ***	0.5441 **	0.4670 **
	(0.1120)	(0.2090)	(0.1578)
	[5.1482]	[2.6026]	[2.9596]
Institutions	5.3393 **	5.0363 *	4.3234
	(2.1663)	(2.8338)	(3.1392)
	[2.4648]	[1.7772]	[1.3772]
Social Inclusion	0.4871 *	0.4594 *	0.3944 *
	(0.2516)	(0.2449)	(0.2084)
	[1.9361]	[1.8764]	[1.8925]
Digital inclusion	0.6380 ***	0.6030 *	0.5177 *
	(0.1075)	(0.3333)	(0.2716)
	[5.9340]	[1.8092]	[1.9059]
Digital inclusion * CO2	0.1547 *	0.1459 **	0.1253 **
	(0.0795)	(0.0673)	(0.0508)
	[1.9468]	[2.1669]	[2.4642]
CO2	0.1898	0.1791 ***	0.1537 ***
	(0.4024)	(0.0192)	(0.0145)
	[0.4717]	[9.3422]	[10.6237]
Trade Openness	0.0453 **	0.0428 **	0.0367 **
	(0.0161)	(0.0157)	(0.0186)
	[2.8088]	[2.7222]	[1.9727]
Investment	0.4535 **	0.4277 **	0.3672 *
	(0.2022)	(0.1968)	(0.1940)
	[2.2428]	[2.1736]	[1.8926]
Education	0.4099	0.3867 **	0.3319 **
	(0.5882)	(0.1724)	(0.1302)
	[0.6970]	[2.2425]	[2.5501]
Inflation	−0.2625	−0.2476	−0.2126 *
	(0.1725)	(0.1679)	(0.1168)
	[−1.5218]	[−1.4749]	[−1.8208]
Constant	36.254 ***	34.196 ***	29.356 ***
	(5.1846)	(5.0460)	(3.8092)
	[6.9927]	[6.7770]	[7.7066]
Diagnostics
Wald test (p value)	3842.5 *** (0.0000)	2154.3 *** (0.0000)	28.743 *** (0.0000)
Arellano–Bond test AR(2) (p value)	1.8472 (0.8993)	0.8362 (0.3421)	1.9876 (0.1174)
Hansen Test for IV (p value)	1.8573 (0.4756)	0.5735 (0.8841)	1.4523 (0.1011)

Notes: The symbols *, **, and *** represent significance at the 10%, 5%, and 1% levels, respectively. The values in parentheses are standard errors and the values in square brackets are the t-values. Note: IV in Hansen test is Instrument Validity.

tabulates the determinants of inclusive growth and its impact on inclusive growth while taking interactions between digital inclusion and carbon dioxide emissions into account. For the high-income countries, digital inclusion has a positive and significant impact on inclusive growth, while carbon dioxide emissions have an insignificant but positive impact on inclusive growth. It is important to mention here that these results explain that industrial development, which comes as the cost of increased carbon dioxide emission, may contribute to inclusive growth; this is undesirable for sustainable inclusive growth. Interestingly, the interaction term between both generates a significant impact on inclusive growth. This outcome highlights the contribution of digital inclusion in making people aware of environmental degradation, which may make inclusive growth a viable proposition. Considering the case of middle- and low-income countries, carbon dioxide emissions, digital inclusion, and their interaction have a positive and significant impact on inclusive growth. People in high-income countries are already highly digitally included, and hence, the impact of digital inclusion is not a significant contributor to inclusive growth.

Table 4. Results for the impact of CO₂ on inclusive growth with institutions * CO₂.

	High Income	Middle Income	Low Income
Inclusive Growth_1	0.4705 ***	0.5441 **	0.4670 **
	(0.0906)	(0.2090)	(0.2178)
	[5.1925]	[2.6026]	[2.1443]
Institutions	4.2259 ***	4.0363 *	4.3234
	(0.9014)	(2.3378)	(3.1392)
	[4.6882]	[1.7265]	[1.3772]
Social Inclusion	0.3993 *	0.4594 **	0.3944 *
	(0.2131)	(0.2149)	(0.2084)
	[1.8736]	[2.1384]	[1.8925]
Digital inclusion	0.5202 **	0.6030 **	0.5177
	(0.2570)	(0.2333)	(0.7161)
	[2.0242]	[2.5847]	[0.7229]
CO2	0.1631	0.1791 *	0.1537 ***
	(0.8285)	(0.0917)	(0.0145)
	[0.1969]	[1.9534]	[10.623]
Institutions * CO2	5.4106 ***	5.6239 **	5.1660 **
	(0.7006)	(2.7490)	(2.3886)
	[7.7222]	[2.0458]	[2.1628]
Trade Openness	0.4833 **	0.0428 ***	0.3672 **
	(0.2058)	(0.0117)	(0.1861)
	[2.3485]	[3.6519]	[1.9727]
Investment	0.2373 *	0.4277 **	0.3589 *
	(0.1330)	(0.1998)	(0.1940)
	[1.7833]	[2.1410]	[1.8497]
Education	0.3312	0.3867 ***	0.4319 ***
	(0.4848)	(0.1253)	(0.1302)
	[0.6832]	[3.0854]	[3.3184]
Inflation	−0.2209	−0.2476 *	−0.2126 **
	(0.1493)	(0.1279)	(0.1068)
	[−1.4793]	[−1.9361]	[−1.9914]
Constant	28.116 ***	34.196 ***	29.356 ***
	(3.0149)	(5.0460)	(3.8092)
	[9.3259]	[6.7770]	[7.7066]
Diagnostics
Wald test (p value)	3813.7 *** (0.0000)	2184.4 *** (0.0000)	3956.1 *** (0.0000)
Arellano–Bond test AR(2) (p value)	2.5735 (0.1164)	1.9342 (0.6244)	1.4784 (0.5483)
Hansen Test for IV (p value)	2.5893 (0.9203)	2.7435 (0.5463)	0.9058 (0.6122)

Notes: The symbols *, **, and *** represent significance at the 10%, 5%, and 1% levels, respectively. The values in parentheses are standard errors and the values in square brackets are the t-values.

below exhibits the interaction between the quality of institutions and carbon dioxide emissions, and what this means for inclusive growth. Institutions are positive and constitute a significant contributor to inclusive growth in high-income and middle-income countries, since institutions are stronger in these two groups. In contrast, carbon dioxide emissions are significant in middle-income and low-income nations, alone. Referring to the latter, environmental degradation policies are less effective and hardly ever enforced. However, the interaction between institutions and carbon dioxide emissions is significant in all three country groups. The results predict that having stronger institutions will help countries in reducing carbon dioxide emissions, which, in turn, will lead to inclusive growth in all three cases [94,98].

Regarding the addition of trade as a control variable in all the tables presented in Section 5, in Table 1, Table 2, Table 3 and Table 4, the role of trade is significant and positive in most cases. A dramatic increase in trade coincides with an equally sharp increase in inclusive growth as trade leads to a decline in extreme poverty and income inequalities worldwide. Trade also increases productive employment opportunities, stimulating economic growth and driving productivity increase. Similarly, inflation was also added as a control variable, and shows a negative impact on inclusive growth in all cases (See Table 1, Table 2, Table 3 and Table 4). Not only are poverty and income inequalities, being important constituents of inclusive growth, the result of the existing social and economic system, but high inflation also doubles down the magnitude of poverty, snatching away people’s purchasing power. Lastly, an important control variable is investment, which plays a positive and mostly significant role in most cases. Investment contributes to the stock of both physical and human capital. Investment also improves the quality and quantity of capital available to an economy and is a crucial determinant of its productivity growth. Investment also contributes to improving education levels, creates jobs, improves health facilities, and decreases poverty and income inequalities, ultimately leading to an increase in inclusive growth.

6. Concluding Remarks

The new mantra in the literature on economic growth is that sustainable inclusive growth must be achieved. This is only possible if environmental degradation is taken care of while considering the pace and patterns of economic development. No matter what policymakers envisage, inclusive growth is still criticized as a mere buzzword for changing the ‘name’ of economic growth. Emphasized in this paper is that inclusive growth is better than other growth models because it takes care of distributional issues while also taking care of environmental degradation, which is often at the cost of development. One of the important issues linked with sustainable inclusive growth, overlooked in earlier literature on economic growth, is the role of environmental degradation. This problem can actually generate productive employment opportunities. Environmental degradation is the outcome of rapid industrial and technological development, which is more problematic in middle-income and high-income countries, since their growth depends on inventions and innovations on a mass scale, regardless of the consequences.

Taking the conceptually fuzzy topic of inclusive growth into consideration, we search for the impacts of digital inclusion, social inclusion, institution quality, and carbon dioxide emissions on countries which are organized into high-, middle- and low-income groups. The interaction of carbon dioxide emissions is generated in three different models with social inclusion, institutional structures, and digital inclusion. Then, their results are compared across high-, middle- and low-income countries. The results predict that all three interactions with carbon dioxide emissions significantly contribute to inclusive growth in all three income groups. Better institutional structures and high levels of social and digital connectivity can help in mitigating carbon dioxide emissions while also embracing sustainable inclusive growth in all three groups. Policymakers should thus be concerned with designing policies that target higher levels of social and digital inclusion. Moreover, improvements in institutional structure should be their focus if they aim to achieve sustainable inclusive growth. In addition, the inflation rate should also be controlled as it can slow down inclusive growth. Trade, education, and investment must be encouraged in all three income groups.

7. Limitations and Scope for Future Research

The index-making exercises show that social inclusion, inclusive growth, institutional quality, and digital inclusion can be decomposed into their key constituents based on strong theoretical pinning. The scope of these indexes can be enhanced by including more descriptive variables into their measurement dependent upon data availability. Long-term inclusive growth is ultimately tied to world population growth, which needs to be catered to in the model, and this exercise naturally raises questions about future predictions of world inclusive growth. This would require future studies to focus on the predictability power of econometric forecasting models that can help in inclusive growth. A plausible conjecture that could explain the cause of hindrance in achieving inclusive growth in future is the increased rates of poverty and income inequalities of the world economy. This explanation suggests that it is possible to achieve a rise in inclusive growth if these two determinants are taken care of at most. To measure their progress, both of these variables have limited data available on them, which needs to be updated at regular intervals of time on a world level. Moreover, for future research, the model may also include the impacts of skilled human capital that can push the technological frontier forward. It is important to recognize that this situation is unsustainable until environmental degradation is considered, which cannot be represented through carbon dioxide emissions alone. Due to world-level data sampling, we were forced to use only carbon dioxide emissions as the proxy for environmental quality. For future analysis, these constraints may be overcome by greater availability of data.