The role of female population, urbanization and trade openness in sustainable environment: The case of carbon dioxide emissions in Ghana

Abstract

Sustainable environment offers many benefits to individuals and societies. As a result, the agenda to reduce carbon dioxide emissions in order to mitigate climate change remains a global concern. Researchers, policymakers and governments have shown interest in this regard. Empirical studies on the subject matter have been increasing conflicting results and little evidence on the effects of some variables necessitate for further studies. To offer value to the literature, in this study, the effect of female population, urbanization and trade openness on carbon dioxide emission in Ghana is assessed. The study used the Stochastic Impacts by Regression on Population, Affluence, & Technology (STIRPAT) model as the foundation for empirical modelling. Time-series data spanning from 1971–2021 were used for regression analysis. In both long run and short run periods, urbanization was noted to exert positive influence on carbon dioxide emission while trade openness and female population exert a negative effect on carbon dioxide emission. Thus, growth in urbanization increases carbon dioxide emission while the opposite effect is the case for trade openness and the female population. Findings from the study suggest the need to intensify the empowerment of women, which could be a crucial catalyst for the achievement of Ghana’s nationally determined contributions toward CO₂ reduction. Also, international trade negotiations that promote environmental protection should not be relaxed.

Keywords:

• gender
• carbon dioxide emissions
• trade openness
• urbanizations

Jel Classification:

• O24
• Q01
• Q53
• Q54
• Q56

1. Introduction

The importance of environmental sustainability has become a topical issue over the last decade. Its importance lies in the promotion of human health, ecosystem, food security, and prevention of the depletion of natural resources. As a global concern, attaining sustainable environment requires committed actions from both developed and developing countries to change their way of life (Smith & Smith, 2020; Udeagha & Breitenbach, 2023c; Udeagha & Muchapondwa, 2022). The pace of forest degradation, climate change, and global warming has heightened concerns for environmental sustainability among world leaders, researchers and practitioners (Alkhidir & Zailani, 2009; Gholami et al., 2016; Wiernik et al., 2013). They are all geared towards achieving a balance among the dimensions of environmental, social, and economic needs (Caniato et al., 2012).

Towards the attainment of sustainable environment, there is consensus to reduce greenhouse gas (GHGs) emissions like carbon dioxide (CO₂), methane (CH₄), and nitrous oxide (N₂O) (Kwakwa et al., 2022; Udeagha & Breitenbach, 2023b). The reason is that these gasses trap heat from the sun and warm the earth/increase the temperature of the earth leading to global warming and climate change (Avishek & De, 2020; Karmellos et al., 2016; Wanapat et al., 2015). Among the GHGs, CO₂ emissions form about two-thirds (Lim et al., 2022). Consequently, this has been the focus of many researchers over the years. World Bank (2021) World Development Indicators show that global CO₂ emissions has increased over the past decade. It saw an increment from 23.3 million kt in 2000 to 29.2 million kt in 2009 and by 2018 it had reached 34.0 million kt. Similarly, carbon emissions (metric tons per capita) was 3.81 and reached 4.60 in 2013. Despite the reduction later to 4.43 in 2016, it has seen an upward trend to 4.48 in 2018. It is worth mentioning that the emission levels have not been equal among developed and developing countries. Although African countries are among the least carbon emitting economies, they are the hardest hit when it comes to the effect of climate change (Alhassan et al., 2019, 2022).

The focus of researchers has thus been to identifying actors that are responsible for CO₂ emissions. It has been documented that factors related to energy consumption (through burning fossil fuels), solid waste, trees (forests), land use changes, industrial processes, certain chemical reactions (namely, cement manufacturing), and transportation indeed give rise to CO₂ emissions (Ansu-Mensah & Kwakwa, 2022, Churchill et al., 2021; Osobajo et al., 2020; Ansu-Mensah & Murad, 2019). Other factors cited in the literature include the usage of electricity, agricultural production and development, industrialization, population explosion, and chemical application in industries (Jiang et al., 2021; Karmellos et al., 2021; Kwakwa, 2021; Kwakwa et al., 2022; Lin & Agyeman, 2019; Liu et al., 2021).

While empirical studies on CO₂ emission abound, urbanization, trade openness, and population effect on CO₂ emissions is gaining attention in recent studies (Aboagye et al., 2020; Ibrahim & Law, 2016; Li et al., 2021; Ngong et al., 2022; Udeagha & Breitenbach, 2023a, 2023d; Udeagha & Ngepah, 2019; Wang et al., 2023, 2023, 2023; Zhang et al., 2017). This may be as a result of increasing pace of population and of urbanization in the world. It is projected that there will be about 8.5 billion people in the world by 2030 and 9.7 billion by 2050 (United Nations, 2022). While over 50% of global population live in urban towns, it is expected that global urban population will hit 6 billion by 2045 (World Bank 2020). These situations are likely to affect the level of CO₂ emissions in many ways. Increased population pressure can increase the loss of forest cover, vehicular traffic congestion, and waste management problems, which can lead to higher CO₂ emissions (Adom et al., 2018). On the other hand, the city compact theory argues that population pressure, especially urbanization, promotes efficiency in the usage of resources and hence can lead to lower CO₂ emissions (Adom et al., 2018). Since the late 1980s, when many developing countries, owing to their subscription to the IMF structural adjustment Programme/Economic recovery programme, were obliged to opened their borders up for more international trade, trade openness has increased among countries. The value of goods and services traded saw a 13% in 2020 and 25% increased to reach $28.5 trillion in 2021 (UNCTAD, 2021). While trade openness can bring about competition to ensure efficiency and hence lower carbon emissions, it can equally trigger acquisition of energy intensive goods which may increase CO₂ emissions (Twerefou et al., 2016).

Ghana’s economy has witnessed some dynamics regarding its urbanization, population, and trade openness. Ghana’s population prior to 2010 was mainly rural. However, since 2010, the number of urban population exceeded rural population. Moreover, the number of urban towns has increased since that period (Adams et al., 2016; Asabere et al., 2020; Songsore, 2020; World Bank, 2021). This has the tendency to increase energy consumption and eventually carbon emissions (Koengkan, 2017). Trade openness has become a key feature of the economy and trading activities. Although international trading activities have increased in Ghana (Solarin et al., 2017; World Bank, 2021), it is expected to increase in the coming years, especially now that the country has become the headquarters of the Africa Continental Free Trade Agreement. Female population continues to dominate male population in Ghana (Tawiah, 2011); however, there has been a reduction in the female share from 51.2% in 2010 to 50.7% in 2021 (Ghana Statistical Service, 2021). The effect of gender on resource conservation has been explored in the literature (Adzawla et al., 2019; Kwakwa et al., 2022; Mensah, 2012; Odonkor & Adams, 2020; Okumah et al., 2021; Kwakwa et al., 2023). However, when it comes to the effect of gender on environmental sustainability, there is scarcity of empirical evidence (Blocker & Eckberg, 1989; Bugri, 2008; Edumadze et al., 2013; Gifford & Nilsson, 2014; McCright, 2010) and CO₂ emissions. It is argued that females are more concerned with the environment and thus protect it more than men (Davis & Fisk, 2014; Hunter et al., 2004; Lee, 2009; Resurrección, 2013; Zelezny et al., 2000). In developing countries, women suffer more when there is a scarcity of resources like water and firewood (Arku, 2015) because of which they are more inclined to engage in activities that would protect the environment than men (Hunter et al., 2004; Okumah et al., 2021).

The rising trend of CO₂ emissions in Ghana as observed over the years (World Bank, 2023) is an issue of concern to policymakers and researchers alike. With the goal of attaining low carbon economy in the coming years via cutting emissions by 64 MtCO₂e by 2030 (UNDP, 2023), appropriate policies, for the Ghanaian economy, emerging from empirical studies are paramount. Following from the above, the main objective of this study is to examine the effect of trade openness, urbanization, and female population on Ghana’s CO₂ emissions. Specifically, the study seeks to assess:

1. the long run effect of trade openness, urbanization, and female population on Ghana’s carbon emission

2. the short run effect of trade openness, urbanization, and female population on Ghana’s carbon emission.

The main question asked in this study therefore is how do urbanization, trade openness, and gender affect carbon emission in Ghana? The questions for the specific objectives are:

1. What is the long run effect of trade openness, urbanization, and female population on Ghana’s carbon emission?

2. What is the short run effect of trade openness, urbanization, and female population on Ghana’s carbon emission?

The contribution of the paper to the literature comes from the angle that it focuses on the effect that the female population has on CO₂ emissions. Thus, although many studies have analyzed the effect of various population dynamics on CO₂ emissions (Kwakwa et al., 2020; Yakubu et al., 2021), it is difficult to come across any on female populationeffect. In the case of Ghana and African countries, this study adds to the limited research that assesses the effect of urbanization and trade. Despite the current urbanization trends and trade dynamics in Ghana, research on their environmental impact is limited compared to other countries (Li et al., 2021; Wang et al., 2023, 2023, 2023).

The remainder of this paper is structured as follows: section 2 is on review of related literature; section 3 presents the methodology, section 4 focuses on the empirical results, section 5 spells out the conclusions and policy implications.

2. Literature Review

The devastating effects associated with GHGs has necessitated researchers and policy makers to unravel their driving forces for policy formulation (Chhabra et al., 2023; Koengkan et al., 2022). Studies on CO₂ emissions have increased over the recent decades. This shows the extent of attention researchers have given towards the attainment of sustainable environment. It has often been argued that economic activities put pressure on environment, thus degrading its quality (Koengkan et al., 2019).

2.1. Trade openness and carbon emission nexus

International trade has been identified as one of the major economic activities that exert influence on the environment (Udeagha & Ngepah, 2022; Wang et al., 2023). This is because international trade relies on energy for the production, exportation, and importation of goods and services (Chhabra et al., 2023). Thus, the expansion of international trade could result in higher CO₂ emissions. Also, trade facilitates households’ accessibility to energy-intensive equipment that may lead to higher pollution. On the other hand, international trade enables countries to import energy-efficient gadgets for households and industrial use, which may lead to lower energy usage and lower CO₂ emissions (Aboagye, 2017; Adom et al., 2018).

Empirically, Leitão (2021) found that international trade reduces CO₂ emissions among European Union countries. Earlier Leitão (2021) had found that trade intensity reduces CO₂ emissions in Portugal. Among ASEAN countries, Pata et al. (2023) established among other things that international trade reduces carbon emissions. However, Charfeddine and Khediri (2016) found trade increases CO₂ for the United Arab Emirates. Vural (2020) in a study obtained a positive effect of trade openness on carbon emissions in sub-Saharan Africa. In Ghana, Abokyi et al. (2021) also found that trade openness increases carbon emissions, but Boateng (2020) found the opposite effect. Chhabra et al. (2023) reported that trade openness aggravates carbon emission for the BRICS economies. For the economy of France, Omri and Saadaoui (2023) estimated a positive effect of trade openness on carbon dioxide emissions. A study by Ashraf et al. (2023) which explored effect of trade openness on CO₂ emissions in 75 BRI countries confirmed trade expansion increases carbon emissions.

Some studies have also reported on mixed results. For example, Anasari et al., (2020) also found a mixed evidence in their study when trade was found to reduce CO₂ emissions in USA and Saudi Arabia but increases emissions in Canada and Saudi Arabia. In the case of Iran, Australia, and Spain, they found insignificant effects. Kim et al. (2019) analyzed the effect of trade and found conflicting results in the sense that trade with the North is associated with an increase in CO₂ emissions, whereas trade with the South reduces CO₂ emissions. Also, advanced countries trading activities with the South or the North reduces carbon emissions, while developing countries trading with the North leads to higher CO₂ emissions, but the opposite is the case when they trade with the South.

2.2. Urbanization and carbon dioxide emissions nexus

Population pressure is noted to affect the level of CO₂ emissions (Liang et al., 2023, 2023). Population growth implies more mouths to feed, people to accommodate, and office spaces to be provided. Consequently, more energy is consumed, which will lead to higher carbon emissions (Lv et al., 2023). Land space also becomes limited due to population pressure. More lands are cleared for farming and accommodation purposes as population of urban towns increases, which reduces forest cover and thereby increases CO₂ emissions (Kwakwa et al., 2022). However, some argue that population pressure like urbanization which is associated with conglomeration of businesses tend to lead to lower carbon emissions due to economies of scale (Adom et al., 2018; Kwakwa & Alhassan, 2018). Regarding population pressure, various studies have looked at the effect of urbanization, population density, and dependency ratio. The reported evidence from many of these studies is that urbanization increases carbon emissions.

For instance, the works of Adusah-Poku (2016) and Boateng (2020) showed urbanization increases CO₂ emissions in Ghana; Boamah et al. (2017) found similar results for China. However, Saidi et al., (2017) found urbanization reduces CO₂ emissions in emerging countries; and Asumadu-Sarkodie and Owusu (2017) found positive effect for Senegal. Chen et al. (2023) obtained a positive effect of urban growth in the carbon emissions in 125 countries. In their study on the effect of urbanization on carbon emissions, Dutta and Hazarika (2023) focused on 68 low-income and lower-middle-income countries and found that urbanization increases carbon emission. Wei et al. (2023) also found for the belt and road initiative region that urbanization positively increases carbon emissions. A comparative study on the carbon effect of urbanization for South and East Asia by Amin and Song (2023) revealed urbanization increases carbon emissions of East Asia regions but insignificant effect in South Asia region.

However, Li and Haneklaus (2022) in assessing the drivers of carbon emissions in G7 countries found urbanization pressure to reduce carbon emissions. In BRICS economies, the work of Balsalobre‐Lorente et al. (2022) found urbanization to be helpful in the reduction of carbon emissions. Earlier studies like Sharma (2011) and Hossain (2011) found urbanization reduces carbon emissions for a group of developing and developed economies. Also, Ahmed et al. (2019) found similar outcome in Indonesia.

2.3. Gender and carbon emissions

The effect of gender of CO₂ emissions has not been much investigated. Women are argued to be concerned with sustainable environment than men since the former suffer the most from environmental degradation (Kwakwa et al., 2022). Women are therefore noted to engage in conservation practices compared to men (Khan, 2023). Studies like Khan (2023), Ghasemi et al. (2021) and Uliczka et al. (2004) among others have also confirmed that women engage in activities that positively affect environmental management. For instance, in the USA, Atif et al. (2021) reported that female board members promote renewable energy consumption. Regarding the gender effect on CO₂ emissions, few studies such as Nuber & Velte (2021) reported that female board members promote efforts to reduce total carbon emission by firms in Europe. Fan et al. (2023) found that female directors outside of a firm help to reduce firms’ carbon emission in Japan.

2.4. Gaps in the literature

The above review has shown that studies on carbon emissions has gained the attention of researchers. However, few studies have analyzed the effect of trade and urbanization in Ghana with reported conflicting results. Also, the effect of female population on national CO₂ emissions has not been given much empirical analysis. As it stands now, the few studies have focused on how gender affects firm level emissions with little to no evidence at the national level. Such studies are also outside Africa, which makes it difficult to inform policymakers on the continent for the appropriate actions to be taken. These issues make it necessary for the current study to be embark upon for the Ghanaian economy at this moment.

3. Methods

3.1. Theoretical and empirical modeling

The concept of sustainable development since 1987 has been used to campaign for societies to meet their current needs without compromising the needs of the future generations. Thus, it encompasses three main pillars, namely, the economy, society, and environment. The environment pillar focuses on protecting the environment from being destroyed by human activities geared toward meeting current needs. The idea behind this pillar is the need to properly manage human activities that put pressure on the environment in order not to disadvantage future generations. Accordingly, the Stochastic Impacts by Regression on Population, Affluence, & Technology (STIRPAT) model was been developed to assess which factors may have environmental effects (Dietz & Rosa, 1997). The model argues environmental impact (I) is attributed to population (P), affluence or economic activities (A), and technology (T). This is expressed in a stochastic form as:

(1) $I\hspace{0.17em}=\hspace{0.17em}a.P^{\beta }.\hspace{0.17em}{A}^{\lambda }.T^{\sigma }.v$(1)

The variables a, β, λ, σ, δ, and v are parameters to be estimated. Since CO₂ emission is a major contributor of GHGs, it is considered as an environmental problem (Minlah & Zhang, 2021), carbon emissions are taken as Impact (I). Following the works of Xu et al. (2020) and Dietz and Rosa (1997), female population (FEP) and urban population (UBS) are used to denote (P) population pressure. Affluence is denoted by trade openness (TO), which contributes significantly to income level of Ghana. Trade openness has also been associated with the wealth of a nation (Froyen, 2013) and telephone infrastructure is used to represent technology (TEL) (Kwakwa et al., 2022). The choice for this model is appropriate for this study, taking into consideration the recent population dynamics in Ghana, carbon emissions level, and the level of international trade in the country.

Taking the above into consideration gives:

(2) $C{O}_2=\hspace{0.17em}a.FEP{O}^{\beta }.UB{S}^{\lambda }.T{O}^{\sigma }.TE{L}^{\delta }.v$(2)

The natural logarithm of equation 2 is taken to transform the equation to:

(3) $LC{O}_{2t}=\alpha \hspace{0.17em}+\beta LFEP{O}_t+\hspace{0.17em}\lambda LUB{S}_t+\sigma LT{O}_t+\hspace{0.17em}\delta LTEL\hspace{0.17em}+\hspace{0.17em}{\vartheta }_t$(3)

3.2. Data and estimation techniques

Time series data spanning from 1971–2021 is used for analysis. World Bank’s (2023) World Development Indicators is the source for all the data used in this study. First of all, the variables for the study are chosen based on the objective, theoretical, and empirical review. The period 1971–2021 is chosen due to data availability for the chosen variables and the fact that it is in the relevant time frame to inform policymakers. CO₂ emission is measured by CO₂ emissions metric tons per capita, female population is share of female in total population, urbanization is urban population as a share of total population, trade is measured as the sum of exports and imports as a share of GDP, and telecommunication infrastructure is measured as the number of telephone subscribers. The use of these proxies does not deviate from what previous studies have used (Aboagye et al., 2020; Xu et al., 2020). Time series data usually contain unit root which can lead to spurious results. To avoid this, it is important to ascertain the stationarity property of the variables. If a variable is stationary, it is free from unit root. If a variable is not stationary at levels, the first difference is taken to verify if stationarity is confirmed. This study used the Augmented Dickey-Fuller (ADF), Phillips-Perron (PP), and Zivot-Andrew (ZA) unit root tests to assess the stationarity property of the variables. The ADF and PP are appropriate for sample size used for this study. However, if a variable contains a structural break, they may not give credible results. Consequently, the ZA test, which is effective in the presence of structural breaks, is also employed to confirm the results (Zivot & Andrews, 1992).

Then, we tested for the long run relationships among the variables using the Autoregressive Distributed Lag (ARDL) Bounds approach, which is ideal for a mixture of variables integrated of order zero I (0) and one I (1) (Pesaran & Shin, 1995). Having established the long-run relationship, the study followed with ARDL regression to estimate equation 4. This technique uses the lags of both dependent and explanatory variables for regressors. The maximum lag selected was 4 based on the Akaike information criterion. The choice for this estimation method is predicated on the fact that it is ideal for cointegrated variables of different orders of integration and is able to address the potential endogeneities associated with the models unlike estimators like the ordinary least squares technique. The ARDL framework of equations (4) can be presented as respectively follows:

(4) $LC{O}_{2t}=\alpha \hspace{0.17em}+\beta LFE{P}_t+\hspace{0.17em}\lambda LUB{S}_t+\sigma LT{O}_t+\hspace{0.17em}\delta LTEL\hspace{0.17em}+\hspace{0.17em}{\vartheta }_t$(4)

(5) $\begin{array}{rl}& \mathrm{\Delta }LC{O}_{2t}=\hspace{0.17em}{\rho }_1+\hspace{0.17em}\sum {\pi }_1\mathrm{\Delta }LC{O}_{t-i}+\hspace{0.17em}\sum {\beta }_1\mathrm{\Delta }LFE{P}_{t-i}+\hspace{0.17em}\sum {\lambda }_1\mathrm{\Delta }LUB{S}_{t-i}+\hspace{0.17em}\sum {\mathrm{\Omega }}_1\mathrm{\Delta }LT{O}_{t-i}+\hspace{0.17em}\sum {\eta }_1\mathrm{\Delta }LTE{L}_{t-i}\\ & +\hspace{0.17em}{\phi }_{11}LC{O}_{t-i}+\hspace{0.17em}{\phi }_{12}LFE{P}_{t-i}+\hspace{0.17em}{\phi }_{13}UB{S}_{t-i}+\hspace{0.17em}{\phi }_{14}LT{O}_{t-i}+\hspace{0.17em}{\phi }_{15}LTE{L}_{t-i}+{\upsilon }_{2}ec{m}_{t-1}+\hspace{0.17em}{e}_{1t}\end{array}$(5)

where ρ_{1 is} drift components and e_1t is the white noise error term. The terms with summation signs (Ʃ) represent the short-run dynamics, and the second part of the equation with the coefficient φ_ij corresponds to the long-run relationship. The ecm_t-1 is the error correction term, with coefficient ʋi referring to the speed of adjustment to long-run equilibrium and ∆ is a difference operator. After the estimation of the short and long-run coefficients of the explanatory variables, diagnostic tests of normality, heteroscedasticity, serial correlation, and stability were conducted to establish the adequacy of the model.

The Fully Modified Ordinary Least Squares (FMOLS) and the Canocical Cointegration Regression (CCR) were estimated to confirm the long run results from ARDL technique. Also, the variance decomposition analysis to ascertain the contributions of the drivers of CO₂ emission following a shock is also analyzed. For causality analysis, Toda Yamamoto causality test is employed. The study relied on Eviews version 10 for the data analysis. It was used for the preliminary tests and the model estimations. Following Fuinhas et al. (2021) and Kwakwa et al. (2022), the methodological approach has been illustrated in Figure 1.

Figure 1. Estimation technique flow chart.

4. Results and discussions

4.1. Unit root and cointegration results

Unit root tests from the ADF, PP, and Zivot-Andrews approach are Table 1. All three tests confirm that trade openness, telecommunication, and CO₂ emission are stationary at first difference. Also, the ADF and PP do not confirm stationarity at levels and first difference for female population and urbanization; ZA reveals that they are at stationary at levels. These results indeed show in the presence of structural breaks the ADF and PP may give misleading results. The attainment of stationarity of the variables at levels and first difference implies the variables can be used for regression analysis without getting a spurious outcome. In Table 2, cointegration results using the ARDL approach are reported. The results reveal cointegration exists between carbon emission, female population, urbanization, trade openness, and telecommunication. Thus, the female population, urbanization, trade openness, and telecommunication can be factors contributing to the level of CO₂ emissions in Ghana.

Table 1. Stationarity tests results

	ADF, t-statistic		PP, t-statistic		ZA, t-statistic
Series	At levels	First diff.	At levels	First diff.	At levels	First diff.
LTO	−1.3824	−5.8695***	−1.3864	−5.7995***	−3.2547 (1986)	−7.1424*** (1983)
LFEPO	−2.3353	−0.3106	−1.3822	−1.933	−7.5657*** (1984)	-
LUBS	−3.6128		0.6900		−4.2510*** (1985)	-
LTEL	−0.6824	−7.4525***	−0.8174	8.2845***	−4.1203 (1997)	−9.287*** (2008)
LCO2	−0.6453	−7.5495***	−0.4531	−8.606***	−3.9737 (1984)	−9.356*** (1992)

***means significant at 1% level

Table 2. Results for cointegration test

Test Statistic/significance	Value	K
F-statistic	2.916631	4
Significance	Lower Bound	Upper Bound
10%	2.2	3.09
5%	2.56	3.49
1%	3.29	4.37

4.2. Regression results

The regression results reported in Table 3 show that in the long run, urbanization, female population, and trade openness significantly determine the level of CO₂ emissions in Ghana. In particular, it is seen that trade openness has statistically significant negative coefficients ranging 0.343–0.747 depending on the estimator. This demonstrates that a 1% expansion towards greater openness to trade contribute to declines in CO₂ emissions by nearly 0.34–0.75%. Thus, the effect of greater openness on CO₂ emission is favourable to the global fight CO₂ emissions. Similarly, depending on the estimation procedure, female population has statistically significant negative coefficients between 0.434–10.45, implying that there is about up to 10% reduction in CO₂ emissions per every 1% increase in the country’s female population. In contrast, urbanization has statistically significant positive coefficients ranging 9.085–19.91. This is an indication that as urban population increases by 1% the corresponding rise in CO₂ emission is between 9–20%. The short run dynamics is estimated using the ARDL (see Table 4). The results indicate that, in the short run too, both trade openness and female population have negative statistically significant effect on CO₂ while the contrary is established for urbanization. Specifically, a 1% expansion in trade openness and female population is respectively accompanied by about 0.27% and 4.7% decline in CO₂ emissions, while 1% rise in urban population is associated with nearly 8.9% increase in the level of CO₂ emissions in the short run. Both the short and long run results show that trade openness and female population are important aspects of the drive to reduce CO₂ emission in Ghana, but urbanization worsens it.

Table 3. Long run regression results

	ARDL RESULTS		FMOLS RESULTS		CCR RESULTS
Variable	Coefficient	Std. Error	Coefficient	Std. Error	Coefficient	Std. Error
LTO	−6.1926***	2.2720	−3.379***	1.0631	−3.379***	0.9172
LFEPO	0.366***	0.2795	0.0423***	0.1249	0.0357***	0.1321
LUBS	−0.7740***	0.2718	−0.4427***	0.1067	−0.4391***	0.1220
LTELE	11.15345	3.7933	7.7543	2.0921	7.782	1.8690
C	57.469**	23.993	25.46198**	10.1862	25.422***	8.6795
Adj-R^2			0.73		0.72

*** and ** respectively means significant at 1% and 5% level

Table 4. ARDL short run regression results

Variable	Coefficient	Std. Error	t-Statistic	Prob.
D(LFPO)	−22.587415	12.819196	−1.761999	0.0855
D(LTEL)	0.101812	0.072384	1.406547	0.1671
D(LTO)	−0.215211	0.070616	−3.047610	0.0040
D(LUB)	3.101088	1.054821	2.939919	0.0054
ECM(−1)	−0.278038	0.087388	−3.181654	0.0028

In Table 5, diagnostic test results for the regression analysis are reported. It reveals that the regression results do not violate the key assumption of the linear regression about the error term. Thus, normality, serial correlation, and heteroskedasticity tests confirm the residual term is normally distributed, and it does not suffer from autocorrelation and heteroskedasticity. Also, the stability test in Table 5 as well as CUSUM and CUSUM of squares shown in Figures 2 and 3, respectively, confirm the model is stable over the period. These diagnostic tests therefore reveal the robustness of the model.

Figure 2. CUSUM test for stability.

Figure 3. CUSUM of squares test for stability..

Table 5. Diagnostics results

Diagnostic	Test	Statistic	Probability
Normality	Jarque-Bera	0.59	0.75
Serial correlation	Breusch-Godfrey	2.0	0.10
Heteroskedasticity	Harvey	1.16	0.35
Stability	Ramsey RESET	1.25	0.21

4.3. Causality and variance decomposition analysis

Table 6 shows the causal relationships among the variables. Urbanization and trade openness granger causes carbon emission; urbanization and female population granger cause trade; female population, telecommunication, and carbon emission granger cause urbanization; and urbanization and telecommunication affect female population. It does suggest that there is a bidirectional relationship between trade and carbon emission; carbon emission and urbanization; and trade and female population. A unidirectional causality is also noted to move from trade to telecommunication. It is important to highlight that evidence of Granger causality does not necessarily establish causal inference achieved through instrumentation. To assess the actual contribution of urbanization, female population, telecommunication, and trade openness to CO₂ emissions over the years, the variance decomposition analysis was done using the Cholesky decomposition. Table 7 shows that over a 20-year period, all variables increase their share of the impacts. Urbanization throughout the period exerts the greatest influence, followed by female population, trade openness, and infrastructure. At period 2, the share of urbanization, female population, trade, and telecommunication was 4.7%, 13.3%, 0.34%, and 0.40%, respectively. Urbanization’s influence then increases to 17.4% in period 5, reduces to 15.8% in period 10, and then to 14.8% in the 20^th period. Within the same period, female population also sees its effect increase to 23.6% in period 5, reduction to 22.9% in period 10 and then 21.0% in period 20. Within the same period, the share of trade stood at 2.5% in period 5, increased to 10.46% in period 10, and then to 10.40% in the 20^th period.

Table 6. Causality results

Dependent Variables
	LCO2	LTO	LTEL	LUBS	LFEPO
Covariates	Chi-sq
LCO2		4.49	0.57	3.26	1.57
LTO	11.82**		2.89	11.02***	23.00***
LTEL	0.070	0.921		0.57	2.443
LUBS	7.483**	32.79***	0.25		2.90
LFEPO	0.05	30.19**	1.64	4.876*

***; ** and * respectively means significant at 1%; 5% and 10% level

Table 7. Variance decomposition results

Period	S.E.	LCO2	LTO	LUB	LTEL	LFEPO
1	0.096863	100.0000	0.000000	0.000000	0.000000	0.000000
2	0.110081	81.16993	0.349193	4.766239	0.402880	13.31176
3	0.124910	63.25683	1.287345	14.75916	0.980877	19.71579
4	0.130681	57.94271	1.329898	17.58315	0.905150	22.23909
5	0.134374	54.90154	2.565189	17.43273	1.441873	23.65868
6	0.140416	50.30356	5.928071	16.92748	2.763660	24.07723
7	0.146796	46.29971	8.095442	17.14761	4.498614	23.95863
8	0.151678	43.64928	9.078085	17.13730	6.501978	23.63335
9	0.155418	41.75392	9.816368	16.54108	8.605090	23.28354
10	0.158817	40.16839	10.46652	15.84479	10.54872	22.97159
11	0.161881	38.92349	10.83026	15.26157	12.21923	22.76546
12	0.164511	38.00745	10.90264	14.83111	13.64129	22.61751
13	0.166827	37.27435	10.84171	14.58334	14.85049	22.45011
14	0.168969	36.63280	10.74888	14.51221	15.85953	22.24658
15	0.170940	36.08671	10.65460	14.53966	16.68863	22.03040
16	0.172679	35.64930	10.56379	14.59750	17.37127	21.81814
17	0.174163	35.30087	10.48485	14.66511	17.93697	21.61220
18	0.175406	35.01621	10.42932	14.73274	18.40493	21.41679
19	0.176424	34.78475	10.40292	14.78301	18.78869	21.24063
20	0.177230	34.60223	10.40320	14.80409	19.10020	21.09029

5. Discussions

The results emanating from the regression and causality analyses are unambiguous evidence that Ghana’s greater openness to trade exerts downward pressure on CO₂ emissions. Since trade openness comprise exports and imports, then greater trade openness and for that matter declining CO₂ emissions implies a rise in either one of these components or both. Ghana’s exports emanate largely form the extractive and agricultural sectors such as oil, gold, cocoa etc., while imports are widely diversified including, but not limited to, daily consumables, durables, automobiles, technology etc. The very nature of the country’s export composition has the potential to put upward pressure on CO₂ emission, but the overall trade openness-CO₂ emission nexus established by the study shows that this may have been entirely counteracted by its import composition to the extent that the net effect results in decline in CO₂ emissions. The results could also indicate that trade openness has enabled Ghanaian economy to benefit from the importation of energy efficient technologies for industrial and domestic purposes. For instance, even though Ghana’s trade regulations are generally less stringent, government has been implementing regulations on the important and use of efficient appliances such as air conditioners, refrigerators, and light bulbs (Energy Commission, 2018). This could underpin the CO₂ emission reduction gains associated with Ghana’s greater trade openness because refrigerator for instance is the second most used appliances by consumers accounting for a significant share of electricity consumption (Agyarko, 2016). Prior to the regulation there were over 6 million refrigerators used by Ghanaian households, over 2 million of which were inefficient (Gyamfi et al., 2018) with each believed to generate over 0.7 tons of carbon emissions annually (Gyamfi et al., 2018). The regulation applies not only to households but also to industries where government is even targeting a 20% double improvement in the country’s industrial energy efficiency by 2030 as part of its nationally determined contributions to the global fight against climate change. Thus, the outcome of this and allied trade regulation is concurrent increase in openness and a decline in CO₂ (Asinyaka, 2019). Moreover, recent efforts by government to attain low carbon economy could have triggered trade openness to reduce carbon dioxide emission. Previous studies in Ghana including Aboagye (2017) reported a positive effect of trade on carbon dioxide emission. Also, Wang et al. (2023) found trade openness reduces carbon emissions for high income countries but it increases emissions in poor countries. The differences in the results could be attributed to the variations in sample period because the study had relied on data ex-ante the regulation. Then again, the recent trade environment related policies could be the reason behind the carbon emission reduction effect noted in this study. Kwakwa et al. (2020), a relatively recent study also observed this negative effect though not statistically significant. At the regional level and Asia Ibrahim and Law (2016) and Zhang et al. (2017) have shown showed trade openness reduces carbon dioxide emissions in sub-Saharan Africa and China respectively.

Like greater openness, it is also observed that an increase in female population is associated with a decline in the level of CO₂ emissions. This is evident from both the regression and Granger causality analysis. Indeed, several studies have shown that women are more inclined to demonstrate resource and environmental conservation relative to men (Buckingham, 2010; Dietz et al., 2002; Kwakwa et al., 2022). For instance, many females prefer to use cleaner energy sources to meet their domestic fuel needs since they tend to be adversely affected by the use of fuelwood and other forms of traditional fuels (Das et al., 2014). The use of fossil-fueled cars in Ghana is an active source of CO₂ emission, and there are less women that own and drive cars in the country could partially account for the fall in CO₂ emission following an increase in female population. Again, women in the country seem to travel more in public transport than men while those women who own private vehicles, arguably, embark on less trips than men. Also, females generally often tend to spend more on low-emitting products such as clothing, furnishing, and healthcare. Women are also found to consume less processed meat and beverages, all of which contribute to lower carbon emission. Based on the above, it is not unreasonable to expect the level of CO₂ emissions to reduce as the country’s population becomes increasingly female dominant. At the household level, studies like Kwakwa and Adu (2015) have reported that female households engage in pro-environmental behavior. Although Ghana’s gender population gap has been closing in recent times, female population remains dominant. The gender structure implies that the country could harness its female population to achieve environmental sustainability, especially through reduced CO₂. Similar findings are observed in other countries by Buckingham (2010) and Dietz et al. (2002). These studies argue, among other things, that women are generally more active in environmental reform projects and tend to perceive environmental risks as more threatening. This is in line with the growing body of literature showing how women manage resources, consume modern energy, and prioritize climate change differently than men.

The harmful effect of Ghana’s urbanization on the level of CO₂ emission could emanate from a number of channels. Foremost, the process of urbanization in Ghana is typically a result of rural-urban migration. This puts excess pressure on urban towns and translates into heightened demand for housing and other residential and socio-economic infrastructure (World Bank, 2022). Clearly, these are avenues for non-negligible loss of forest covers which could have absorbed CO₂ emissions. Secondly, it must be acknowledged that Ghana’s urbanization is frequently a by-product of industrialization. Indeed, unlike in other countries, Ghana’s urbanization is characterized by mass industrial activities clustered in the same few urban cities since they tend to be the hub of manufacturing firms resulting from the availability of reliable power supply and ready market. The activities of construction and industry imply substantial resource and material use which could contribute to CO₂ emission. Thirdly, and a more active source of CO₂ emission associated with urbanization in Ghana relates to the use of energy which is primarily a fossil. In addition to the above, the traffic congestion in Ghana’s urban towns creates situations where more fuel is burned which contributes, although in indirect ways, to the higher level of CO₂ emissions in urban towns of Ghana. Again, the slum conditions and waste management problems could lead to higher carbon dioxide emissions. The combined effects of these channels are a rise in carbon dioxide emissions. Thus, for Ghana like many other developing countries, urbanization means more forest loss, material use, waste, fossil energy consumption with each contributing to increases in carbon emissions. This is consistent with the evidence that cities are responsible for close to 70% of global CO₂ emissions associated with energy consumption. The findings fall in line with Kwakwa and Alhassan (2018) who found urbanization in Ghana to intensify CO₂ emissions. Comparable findings are reported by Yuzhe et al., (2016) for China, Wang et al. (2019) for China, Anser et al. (2020) for SAARC countries and Musah et al. (2021) for West African countries.

6. Conclusion, policy implications, limitations and further studies

6.1. Conclusion

The attainment of reduced CO₂ emissions remains a critical concern to both developing and developed countries. Researchers have therefore investigated the drivers of carbon dioxide emission. Since several concerns exist still in this area, , this study investigated the effect of trade openness, female population, and urbanization on Ghana’s CO₂ emissions. Time series sourced from the World Bank for the period 1971–2021 was used for this study. Data were analyzed using regression and variance decomposition analysis. It emerged from regression analysis that greater trade openness translates into lower CO₂ emissions; female population also reduces CO₂ emission while urbanization increases CO₂ emission. A Variance Decomposition Analysis showed the greatest effect comes from urbanization over a 20-year period. This was followed by trade openness and female population. In recent decades, Ghana’s trade is much focused on increasing the attractiveness and diversification of its exports. This has culminated into efforts of shifting away gradually though steadily from the export of raw natural resources to processed natural resources. Exports of non-traditional agricultural products are also becoming dominant in the country’s trade agenda. Imports remain generally diversified and currently entailing more technology and innovation, which are, in turn, deployed to support domestic production and eventually exports. Without doubt, urbanization has become part of Ghana’s economy. This has been occasioned mainly by the high spate rural-urban migration mainly because of the quest of seeking for greener pastures in the urban areas. Since it may be difficult to deurbanize Ghana, measures should be put in place to reduce pace of growth in urban population as well the associated environmental pressures. The gender population gap in the country is almost closed as there are more females as males with females accounting for over 50% of Ghana’s population. However, several gender inequalities still persist in many dimensions such as education, healthcare, representation as well as involvement in critical environmental and climate change decisions. Yet, this study has shown evidence consistent with recent the Paris Agreement that improving gender inequities is important to any meaningful discussion of the intersection of female empowerment and climate change. This is considered as an important contribution of the study and moderately novel in the Ghanaian context. This is because studies on CO₂ emissions in the country have entirely neglected the active role and participation of female population in its exclusivity. Indeed, as highlighted in the introduction, although many studies have analyzed the effect of various population dynamics on carbon dioxide emissions in Ghana, the aspect regarding female population appears to be entirely unexamined while that of trade openness and urbanization is scarce.

The contribution of the paper to the literature therefore emanates from its focus on the effect that female population has on carbon dioxide emissions. This is still an important scientifically established empirical contribution that deepens our collective understanding of the role of females/women in the global fight against CO₂ emission. The limited studies on the effect of urbanization and trade openness in Ghana and Africa in general is bridged in this study. These certainly have stern implications for both literature and policy discussions on women, trade openness and urbanization on one side and environmental sustainability, climate change, global warming among other environmental outcomes on the other side.

6.2. Policy implications

The main policy implication emanating from the study relates to the evidence that Ghana can rely on its female population to reduce carbon dioxide emissions. Yet, Ghana, neither has any policy that specifically influences gender population gap nor is it within the realm of policy to feasibly alter reproduction to increase female population. It is for this reason that efforts need to be put in place to enhance the socio-cultural and economic livelihoods of the female population in the country. The country can thus, benefit more from its female population in terms of quality environment by further enhancing their capacity to engage in environmentally friendly practices. In this way, intensifying the empowerment of women through girl child education, female literacy, female labour force participation, and women parliamentary representation could be a crucial catalyst optimizing the environmental gains of its female population. This is because if female population on its own is able to reduce CO₂ emissions, then increasing women empowerment, female literacy, female labour force participation, and women parliamentary representation could be expected to generate even more of such environmental gains. Even though Ghana already has policies aimed at bridging the gender inequality gap in respect of education at all levels supported by various affirmative actions and other female literacy programmes, gender inequities still persist. Similarly, although there are also policies aimed at removing barriers to female labour force participation and increasing women representation at all arms of government (i.e., executive, parliament and the judiciary), more efforts on these fronts are imperative for greater gains. This study recommends intensification of these and other allied policies as they would not only help the country achieve its respective objectives but also contribute to reduction in CO₂ emissions which would in turn support the realization of Ghana’s nationally determined contributions toward global CO₂ reduction.

The findings and the ensuing conclusions also suggest that Ghana can remain on its current trade openness trajectory as part of the country’s broader objective to contribute meaningfully to the global fight against climate change via reduction in CO₂ emissions. To achieve this, we suggest the following in respect of international trade. Foremost, Ghana, generally has a greater trade openness orientation with minimal to moderate protectionist policy. However, the positive environmental gains from trade openness in Ghana should draw the attention of policymakers to remain even more conscious of its trade negotiations with trading partners. In that regard, negotiations that promote environmental protection, particularly in the areas of GHGs such as CO₂ emissions should not be relaxed. It is recommended that regulation on the importation of only energy-efficient households’ appliances such as refrigerators, air conditioners, and light bulbs should be strengthened even further. This should be extended in similar rigorousness to all other sectors of the economy including industry, transport, and construction so that imports that support these sectors are those that are environmentally-friendly and have minimal to zero carbon emission rating. This could be bolstered by increasing import duties on gadgets that are harmful to the environment and/or increase CO₂ emission for carbon-financing projects while imposing lower duties on equipment that leads to lower carbon dioxide emission. In respect of important, government has over the past decade been committed to diversify its export composition as a way to reducing its reliance on the export of natural resource extraction. This is a policy in the right direction because it has the potential of reducing environmental degradation and CO₂ emissions. Policy must not only continue in this pursuit but must also consider enforcing environmental laws and regulation regarding illegal and harmful exploitation and extraction of natural resources that contribute to CO₂ emissions.

Urbanization remains a threat to the country’s efforts to fight hikes in CO₂ emissions as evident by the both regression and causality results. Thus, there is the need for Ghana to re-evaluate its urbanization process and critically in the spirit of environmental sustainability. One viable aspect is to minimize the pace of urbanization which can be achieved when government gets extra committed to rural development. This could balance development in the country and subsequently reduce the rural-urban migration which is mainly responsible for expansion in urban population of the country. Moreover, conscious efforts to reduce transport and congestion by promoting efficient public transport system is crucial. The former is linked to fossil fuels consumption and the latter results in excess waste and fossil use. All of these avenues could contribute to CO₂ emission in both direct and indirect ways. In addition, even though the country has a well-functioning Urban Planning Department coupled with various municipalities and environmental agencies, it appears their effectiveness in planning, directing, and controlling Ghana’s urbanization structure and process to optimize environmental gains is yet to be fully realized. This study recommends a revitalization of the functioning of these institutions to ensure environmentally friendly urban structuring, decongestion, slum degrowth, waste management aimed at reducing CO₂ emissions and improving environmental quality altogether in urban centers. Also, since Ghana’s urbanization process is generally a by-product of its industrial drive, law enforcement agencies need to strictly ensure that industries mainly found in urban areas of Ghana adhere to safer environmental practices.

6.3. Limitations and further studies

An important aspect of the potential effect of trade openness on CO₂ emission relates to the composition and direction of trade as these could have markedly peculiar CO₂ emission outcomes. Similarly, the structure of urbanization could have implications for CO₂ emission different from the effect of the urban population itself. In the same way, some aspects of female population such as female environmental awareness and women’s environmental participation among others could be expected to broaden our understanding of the effects of female population on CO₂ emissions. However, the study was limited in terms of quality data availability and thus could not explore these important dimensions of the theme. This is regarded as a potential source for future studies as quality data become publicly available. Where data are available, future studies can consider various categories of female populations such as the age dimension and employment status to offer more insight into the subject matter.

Author contributions

Paul Adjei Kwakwa conceived the research idea, collected and analyzed data; Peter Ansu-Mensah reviewed literature and concluded the study, Emmanuel OpokuMarfo discussed the results and offered conclusion; and Solomon Aboagye worked on the methodology and offered recommendation. All authors read and approved the final manuscript.

Data availability

The datasets analysed during the current study are available in the World Bank’s World Development Indicators repository, [https://databank.worldbank.org/source/world-development-indicators#]

Informed consent

The study did not rely on data or participants that consent was needed.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Author did not receive any funding for this study