Impact of Environmental Tax on Comparative Advantage of Food and Food Products: A Study of G20 Countries in Light of Environmentally Sensitive Goods

Abstract

Many countries are under constant fear that environmental policies might negatively influence the international competitiveness of polluting industries. In this study, we aim to evaluate the relationship and impact of the environmental tax on comparative advantage of trade in food and food products industry, considered to be one of the highly environmentally sensitive industries. This study also investigates, whether this relationship differs among countries covered in G20, with the help of correlation analysis. We select panel autoregressive distributed lag approach for this study as it can analyse long-run as well as short-run association even when the variables are stationary at different orders of integration. Using panel data from G20 countries over the period of 21 years that is from 1994 to 2015, it is concluded that when we allow environmental taxes to interact with the revealed comparative advantage (RCA) of G20 nations, the overall impact of the environmental tax on the RCA is negative in the long period. It is therefore suggested that countries should follow Porter hypothesis to stimulate innovations resulting from strict environmental regulations that affect the environment in least possible manner.

JEL Codes: C01, C23, C33, F18, O57, Q5

Keywords

Environmental-sensitive goods environmental tax revealed comparative advantage G20 panel ARDL

1. Introduction

With growing international trade, especially in the pollution-intensive sectors, the concern regarding future generations and the environmental degradation is on the rise (Ederington, 2007). Thus, there is need for strict environmental regulations to avoid environmental degradation. But concerns regarding the side effects of such environmental regulations are said to be too expensive to implement (Arnold, 1999) as they might negatively impact variables like employment, income, productivity and overall growth of an economy.

Trade and Environment Debate

The debate linking trade and environmental concerns started around 1970’s (Cagatay & Mihci, 2006) but it became intense near 1990’s with the advent of free trade (Muradian & Martinez, 2001). Stringent environmental policies might affect adversely the international trade of dirty industries influencing their competitiveness and export performance (Kozluk & Timiliotis, 2016). The trade-environment debate can be conceptualised through ‘Pollution Haven Hypothesis’ (Copeland & Taylor, 2004) and Porter hypothesis (PH) (Porter, 1991). Pollution haven hypothesis is mainly grounded on the idea that countries with less stern environmental regulations enjoy comparative advantage in pollution-intensive industries and the countries with tough standards import such environmental sensitive good’s (ESG’s) from them. As a result, the developing countries which gain advantage in dirty exports, soon become pollution havens and end up with less clean industries and more dirty industries.

The other extreme, the PH, suggests that the strict environmental procedures in the host country may result in innovations and lead to increased efficiency for industries located domestically (Porter & van der Linde, 1995). Such innovations resulting from strict environmental guidelines lead to the development of eco-friendly technology, further leading to efficient production and environmental safeguard. The basic idea behind PH is that strict environmental standards compel firms to introduce innovations and adopt green technologies that affect the environment in the least negative manner.

Food and Products—An Environmentally Sensitive Industry

The food and food products manufacturing sector, also referred to by their North American Industry Classification System (NAICS) code (https://www.census.gov/eos/www/naics/), incorporates a vast range of businesses and ever-increasing population ensures great demand. The food and food products industry has an enormous potential that is still unexploited for maximising savings and minimising the cost leading to pollution prevention. This sector uses millions and injects billions into the economy. Food and products accounts for about more than 26% of global emissions of greenhouse gases (GHGs) (Poore & Nemecek, 2018). For most food products, transportation accounts for about 10% in GHG emissions. This manufacturing process involves lethal chemicals, which are released through various channels, polluting the environment. The toxics that are released from manufacturing units are reported in environmental protection agency (EPA’s) Toxic Release Inventory report (US EPA, 2015). It comprises five largest contributors to toxic releases: (a) metal extraction, (b) electrical services, (c) chemical manufacturing, (d) primary metals and (e) paper manufacturing. In comparison to other sectors, this industry faces difficulty in minimising the release of toxic chemicals into the environment. According to toxic release inventory (TRI) report, many food and food products manufacturing industries are aiming to reduce their harmful releases. It should be noted that the TRI P2 tool can prove to be helpful in identifying methods for pollution prevention. Changing the production process, improved cleanliness and ensuring proper scrutiny are some of the ways that can be used to track the production and disposal processes and reduce the releases of toxic chemicals.

The G20 comprises the major players in the global food system. It accounts for 74% of global GHG emissions. Average annual per capita emissions are nearly 7tCO₂e (total CO₂ emissions) (Sebastian, 2015). Australia, Argentina, Canada and France were found to have the highest per capita food-related GHG emissions of G20 countries (foodnavigator.com). Of total carbon emissions for food, the G20 emits approximately 75%. CO₂ emissions account for the principal share of GHGs, which are majorly associated with global warming. Emission intensity, which is the volume of emissions measured in terms of GDP, is better indicator of the country status in terms of pollution. Out of the G20 nations, South Africa has the highest intensity of CO₂ emissions embodied in gross exports of food and food products, and France, being one of the highly advanced nations has the lowest intensity for the same. So far as the absolute figures are concerned, the USA has maximum total CO₂ emissions embodied in gross exports and Mexico has minimum of it (Table 1).

Table 1.

Intensity and Total CO₂ Emissions Embodied in Gross Exports for G20.

Country	Intensity of CO₂ Emissions Embodied in Gross Exports in 2015 (Tonnes per US$ million)	Total CO₂ Emissions Embodied in Gross Exports in 2015 (Tonnes per US$ million)
1. Australia	320.4	5.6
2. Argentina^a	272.9	4.6
3. Brazil	293.5	9.0
4. Canada	338.4	6.5
5. China^a	452.3	21.2
6. India	349.6	6.4
7. Indonesia^a	281.3	7.8
8. Turkey	326.4	3.4
9. EU^a	222.5	26.4
10. USA	309.4	20.9
11. UK	195.9	3.8
12. South Africa	786.3	3.3
13. Saudi Arabia^a	684.6	1.8
14. South Korea^a	371.2	1.9
15. France	193.4	6.6
16. Germany	202.4	9.3
17. Mexico	235.7	2.6
18. Japan^a	220.6	0.9
19. Italy	217.5	5.3
20. Russia^a	617.5	3.8

Source: OECD Statistics.

Note: ^aCountries with RCA < 1.

Environmental Stringency and Environmental Tax

Due to different stringency levels, it, however, becomes more difficult to analyse the stringency and its impact at an international level due to increased complexity that acts as an obstacle for the firms and a source of comparative advantage or disadvantage, depending on the respective country (Silajdzic & Mehic, 2017). Environmental tax is one of the best possible solutions to the problem of environmental degradation due to the expansion of free trade (Gao et al., 2019). Environmental taxes are implemented especially to encourage firms and industries to use the environmental resources efficiently and discourage the use of products and practices that are against the environmental good and also to reduce emission of pollutants to an optimum level and go green. According to the Organization for Economic Co-operation and Development (OECD), tax, the base of which has negative effect on the environment is commonly known as ‘Environmental Tax’. In nutshell, environment-related taxes are mainly imposed for environmental purposes and are considered as an economic tool to address environmental problems. Environmental taxation is one of the best representatives that lead to changes in the relative price of commodities to ensure that polluters perform their activities considering the effects on the environment. However, environmental taxes prove to be more effective and flexible as compared to the command and control measures (Baumol & Oates, 1988).

Food production is one of the industries responsible for a quarter of all GHG emissions, contributing to global warming (Poore & Nemecek, 2019). Manufacturing of food products, especially animal-sourced, is a principal source of carbon dioxide, methane and nitrous oxide emissions, the top three GHGs. The amount of GHGs released by the production of food differs from one food type to the other. CO₂ being the foremost GHG is responsible for emission taxes and tradeable permits for carbon emissions in G20 countries.

If taxes on food products are implemented on the basis of environmental impact of production, then the environmental cost of food manufacturing possibly will be substantially lowered. Further, the amount collected from tax could be used to manage the cost of food products that are more nature friendly (Springmann et al., 2017). It was found that tax on food and food products would lead to 146,000 demises avoided worldwide in the year 2020, two-third of which is due to changes in diet. However, broader tax coverage would increase health benefits as well as raise the tax revenues that could be used to subsidise vegetables and fruits, promoting less environmentally sensitive food products.

II. Literature Review

Environmental Taxes in G20

Pigou (1920) was the first one to propose to reduce environmental pollution by using tax. The environmental taxes on food industries mainly vary by commodity and country due to different managerial and environmental practices. The existing food system is a most important contributor to climate change, causing about 29% of total GHG emissions. Animal-based food products generally have a considerably larger environment impact than plant-based foods (Clune et al., 2017). However, for different emissions levels, taxes on food such as beef, lamb, pork, poultry, vegetable oil, milk, eggs and wheat, rice and for fruits, vegetables, grains, roots, legumes and sugar etc. vary. Variation is also observed with regions (Springmann et al., 2017). The scope of each country’s carbon tax differs, resulting in varying shares of GHG emissions covered by the tax.

Carbon taxes covering 13% of total GHG emissions have already been implemented in some G20 countries, namely Japan, EU, South Africa, Mexico, France and different countries around the globe (World Bank et al., 2016). These carbon taxes cover some of the CO₂ emissions related to food production, especially those emitted from use of electricity and fuels (Moberg et al., 2019).

GHG emissions’ tax in Brazil is disaggregated into different sectors, including food and beverages sector as well. In total, taxes in Italy price 84% of Co₂ emissions from energy use. Italy does not yet have a carbon tax either in place, or under consideration. However, Italy is part of the European Union Emissions Trading System (EU-ETS), which covers 2GtCO₂e of emissions and 45% of the EU’s emissions.

In Russia, tax on food production is a natural tax that is collected in kind (Dmitriev & Novikov, 2018). In 2019, Canada imposed a national carbon tax of $16 a ton of CO₂. New carbon taxes were introduced in Mexico and France in 2013. Australia’s carbon tax, which came into effect in 2012 and is currently set at A$24.15 (US$22.00) per ton of carbon dioxide equivalent (CO₂-e) emitted, covers a broad range of industry sectors and categories of CO₂-e emissions (Robson, 2014). In 2017, Argentina enacted new carbon tax, which taxes the implicit GHG emissions in fossil fuels. As one of the first measures to fight climate change, South Africa introduced a carbon tax. South Africa's carbon emissions are from energy generation and the industrial use of energy and as much as 80% of South Africa's primary energy is powered by coal, affecting production in food sectors (Ntombela et al., 2019). Indonesia does not levy any excise taxes at national level, but a regional tax is imposed on gasoline and diesel, throughout the country. The tax rates on different fuels and uses are linked to Indonesia’s energy use or CO₂ emissions from energy use. Energy use and the CO₂ emissions associated with it include sectors—–transport, industry, agriculture and food, residential and commercial and electricity. Due of Turkey’s limited experience with market-based instruments, considers non-market-based instruments. For example, carbon tax is considered for the sectors to limit GHG emissions, particularly transportation sector.

In the G20, Saudi Arabia has the second-highest CO₂ per capita emissions. In 2000–2010, Saudi Arabia witnessed highest CO₂ emissions in the Arab region. In 2010, it was confirmed that South Korea is considering a carbon tax to help reduce emissions 4% by 2020 (Boden et al., 2011). Germany, like all EU member states, has introduced tax on carbon dioxide emissions from transport and heating which was applied to the numerous fuels—from gasoline and diesel to oil and gas used for heating and fossil fuels like natural gas, diesel, coal and oil used for electricity generation, indirectly playing a massive role in food sector emissions. In 2012, Japan introduced a carbon tax with the goal to take action on mitigating climate change. In 1993, the UK Government introduced, an environmental tax, explicitly designed to reduce carbon dioxide emissions.

According to the Environmental Protection Tax Law of China, two new environmental protection laws were enacted at the start of 2018. One, for emissions, discharge as tax collected from industrial polluters, and the other to fight water pollution.

A Health Association on Climate Change in the UK has implemented a carbon tax that is to be levied on food manufacturers according to the carbon emissions of their produce and ensure reduction of emissions by 2025. It says that food production in UK represents around 20% of GHG emissions.

As a part of GDP, India has one of the lowest environmentally related tax revenue. In India, taxes on energy represented 50% of total environmentally related tax revenue on an economy-wide basis, which significantly contribute to food production. The sectors with the highest environmental tax coverage are agriculture and fisheries (99%) and road transport (97%). Both the sectors play an important role in manufacturing of food and food products. Overall, taxes in India cover 53% of CO₂ emissions from energy usage. Energy use mainly comprises of fossil fuel consumption. According to climate change predictors, the attempt to enforce carbon tax on fossil fuels (petrol, coal and diesel) in India is an attempt to curb CO₂ emissions, which play a vital role in manufacturing of food and related products. According to experts, India charges a carbon tax of US$140 and US$64 per ton on petrol and diesel respectively which primarily comprises of environmental taxes on food production and processing activities. The coal cess that was introduced a few years ago is Rs 400 per ton that is equivalent to 20% of carbon tax.

Models and Methods of Measuring Impacts of Environmental Tax

The stringency of the environmental policies and export competitiveness of firms are considered as one of the long-researched questions (Tewari & Pillai, 2005). A large stream of literature claims to have a negative relation between the two. However, some researchers argue that there happens to be a positive impact of environmental tax on exports. Whereas, some studies even conclude that no significant relationship exists between the two variables.

Various analytical and computational methodologies are used to analyse the impact of environmental taxes on export pattern. Some of these methods include the Leontief input and output model, the Heckscher–Ohlin–Vanek model, the Gravity model, the Vector Autoregressive Models, etc. In this section, an attempt has been made to highlight the models used to study the impact of environmental tax on food and related industries. However, some of the tax-related studies on other products are also mentioned here to understand the various kinds of models and latest invented tools in this area.

General Equilibrium (GE) Model

Matsumoto and Masui (2011) tried to analyse the long-term impacts of carbon tax based on the imputed price of carbon, employing the computable General equilibrium (CGE) model. It considered various energy and non-energy sectors including the food processing sector for almost 15 of all G20 nations and some other nations as well. However, the impact of tax on each country is observed to be different.

Tian et al. (2017) by using a dynamic computable general equilibrium (CGE) model examined food production sector in Shanghai (China) along with other sectors under the tax policies, namely tax24 and tax34. Comparing with the tax44 scenario, the outputs of food production sector decreased by 2.02%, and 0.99% under tax24 and tax34 scenarios.

Ntombela et al. (2019) evaluated the impact of carbon tax policy on food industry, agriculture and other sectors in South Africa, using a dynamic computable general equilibrium model. The agriculture and food sectors show progress in terms of employment and output when carbon tax is put into effect.

Gao et al. (2019) in their research used Dynamic Stochastic General Equilibrium model and observed environmental taxes reduce the export volume.

Partial Equilibrium Model

Rivers and Schaufele (2015) observed that agricultural sector (input to food industry) in British Columbia (Canada) tends to be less carbon intensive than many other highly traded sectors, using Partial equilibrium model and taking carbon tax as an explanatory variable, it concluded that there was an increase in comparative advantage due to the introduction of the carbon tax.

Input–Output Model

Morgenstern et al. (2004) selected the United States for the study of carbon policies impact on manufacturing sectors such as textiles, paper, plastics, food products, leather and chemicals. Using input–output (I/O) analysis they observed that only few industries could stand a burden of carbon tax or any other similar carbon policy.

Bordigoni et al. (2012) employed multi-regional I/O model for analysing the impact of a CO₂ tax levied on food manufacturing and 58 other industrial sectors for European Union countries. They concluded that such carbon tax induces an unnecessary burden on industries and countries.

Mardones and Baeza (2018) estimated the effects of different CO₂ tax rates in three Latin American countries (Brazil, Mexico and Chile) using the Leontief I/O model. The study concluded that environmental taxes decrease the total carbon emissions in these countries.

Regression Analysis

Säll and Gren (2015) made a study estimating the impacts of environmental tax on meat and dairy consumption in manufacturing industries in Sweden. By employing Seemingly Unrelated Regression and constructing Almost Ideal Demand System Model, the results indicated that the introduction of tax on meat and dairy products decrease emissions of pollutants from the livestock sector.

Using multiple panel regression, Cheng et al. (2015) observed that controlling CO₂ emission improved the export competitiveness of Chinese manufacturing industries including the food industry.

Gravity Model

Zhao (2011) investigated a sample of OECD countries from 1992 to 2008. Using Gravity model, it was observed that carbon tax exhibits a negative impact on the competitiveness of energy-intensive industries.

Costantini and Mazzanti (2012) tried to scrutinise the influence of combined environmental taxation on the competitive advantage of manufacturing exports of European Union. Using Gravity model, they concluded that overall environmental policies have positive effect on the competitiveness of European Union.

Silajdzic and Mehic (2017) incorporated an extended Gravity model, and determined that there is a strong positive and significant impact of the environmental taxes on exports of polluting industries in Central and Eastern European countries.

Autoregressive Distributed Lag Technique

Li and Lin (2015) investigated the impact of sales tax on the growth of the US economy during 1960–2013 using the autoregressive distributed lag (ARDL) bounds testing approach of cointegration. It is found that economic growth in the United States is negatively related to sales tax in the long run, but shows positive results in the short run.

Khan et al. (2019) studied the consequences of environmental regulations on carbon emissions in China for the period of 1991–2015, using Non-linear ARDL Technique. The results suggested that there exists a linear relationship between environmental regulations and carbon emissions both in short and long run.

He et al. (2019) studied the environmental performance of environmental tax levied in 36 countries of OECD from 1994 to 2014 based on the Panel ARDL model. It was found that there is a long term cointegration between the environmental tax and the other variables (GHG emissions, nitrogen oxides emissions and sulphur oxides emissions). It was also concluded that the environmental tax negatively impacts nitrogen oxides and sulphur oxides emissions.

Sedehi and Esfahanian (2019) attempted to examine the empirical relationship between gasoline taxes and industrial labour productivity in Iran for the period 1990–2015 using the ARDL approach and the bounds testing to cointegration. The results imply that the tax possess a negative impact on the productivity, irrespective of PH.

III. Data and Methodology

Description About Sample and Variables Under Study

The objective of the present study is to analyse the impact of Environmental tax (ET) on exports of food and food products industry for G20 nations. The G20 group has been selected for the study as it seems to bring together the progressive and emergent economies of the world. Together G20 represent nearby 90% of the world’s GDP, around 80% of worldwide trade, and about two-thirds of the world’s population. Food and food products, one of the top ESG’s, that is comprising of about 10.33% of share in total exports among selected top ESG industries and G20 nations exporting the selected product on an average of about 60% of the world’s product exports (Table 2).

Table 2.

Food and Food Products Export: G20 Countries.

S. No.	Country	1994		2000		2015
S. No.	Country	Absolute (thousand $)	World Percent	Absolute (thousand $)	World Percent	Absolute (thousand $)	World Percent
1	Australia	12895600.84	2.517575	14695981.3	2.2422	34892104.9	1.70677
2	Canada	15744702.22	3.073798	22703015.74	3.463853	59182554.15	2.894952
3	France	50158312.95	9.792278	48694705.05	7.429466	95377244.87	4.665439
4	Germany	34421261.98	6.719974	34247339	5.225197	114954000.1	5.623048
5	Italy	19490982.28	3.805174	22831220.53	3.483413	62261907.14	3.045581
6	Japan^a	3470045.32	0.677448	3538625.481	0.539896	8054182.288	0.393975
7	Brazil	19226294.44	3.7535	19019212.5	2.901806	93321647.5	4.564888
8	Turkey	5505934.76	1.074909	5097149.08	0.777684	23433959.2	1.146287
9	India	4905538.49	0.957696	6401674.29	0.976719	35601472.6	1.74149
10	Indonesia^a	5968322.588	1.16518	6604664.33	1.007689	38246787.1	1.870866
11	China^a	15993243.73	3.12232	18713898.5	2.855224	91022217.4	4.45241
12	Mexico	5375827.5	1.049509	10993158.5	1.677252	36898910.5	1.804934
13	Russia^a	NA	NA	1665867.05	0.254165	20142220.8	0.98527
14	UK	24845489.05	4.850521	26207544.93	3.998547	47914928.39	2.343789
15	USA	68111263.55	13.29719	72696413.83	11.09146	178035168.4	8.708704
16	South Africa	2980199.18	0.581817	3341217.29	0.509777	12012159.4	0.587582
17	EU^a	NA	NA	70516496.7	10.75887	210302825.8	10.2871
18	Saudi Arabia^a	403638.34	0.078801	561824.97	0.085719	5322109.64	0.260334
19	South Korea^a	3434994.97	0.670605	3845102.51	0.586656	11045559.63	0.540301
20	Argentina^a	10396536.82	2.029689	15136888.53	2.309471	47811131.54	2.338712
	G20 total	303328189	59.21798	407512000.1	62.17506	1225833091	59.96241
	Others	208894934.2	40.78202	247914737.6	37.82494	818502972	40.03759
	World	512223123.3	100	655426737.7	100	2044336063	100

Source: Authors’ calculations based on WITS database.

Note: .NA: Not available.

^aCountries with RCA <1.

The principal variable in our study is environmental tax, which points out the sternness of environmental policies and regulations in the exporting country and as mentioned earlier is expected to be the better representative for stringency compared to the other environmental variables like GDP per capita, which are very frequently used variables by researchers. Hence, it is noted that environmental tax forms a better instrument for comparison. Now the question arises ‘how the environmental tax is linked with food industry?’

Food production is one of the industries having the highest possible impact on the environment in terms of GHG emissions (Bonnet et al., 2018). The main GHGs are CO₂, CH₄ (methane) and NO₂ (Piwowar, 2019). Globally, food production accounts for around 25% of all GHG emissions and is almost equivalent to electricity and heat generated worldwide.

According to the Food and Agriculture Organization, food accounts for 30% of energy consumption worldwide, which translates into emissions equivalent to 10 gigatons of CO₂. The food industries uses small coal-fired boilers for heating and pollutants are released due to the lack of pollutant removal facilities. Volatile organic compounds with strong odours are also released from food processing, which impose serious effects on environment and human health (Li et al., 2014). Coal is the most preferred fuel used in countries like India and China for electricity generation and various industrial processes involving heat. Main emissions from coal fired and lignite based manufacturing processes are CO₂, NO _x , SO _x and air-borne inorganic particles such as fly ash, carbonaceous material (soot), suspended particulate matter and other trace gas species (Mittal & Sharma, 2003). Food system activities, including producing food, transporting and storing it, produce GHG emissions that contribute to climate change.

Environmental tax, an indicator of environmental policy stringency index, includes CO₂ tax, diesel tax, NO _X tax and SO _X tax (OECD Statistics). As discussed above, food industries emit CO₂, NO _x , SO_x, etc. during manufacturing and at different channels of supply chain. Thus environmental tax can be used as a proxy to study the impact of environmental policy stringency on the comparative advantage of food and products.

Export potential of food and food products industry is represented by revealed comparative advantage (RCA) in the present study. The main data source for the present study is OECD Statistics and World Integrated Trade Solution (WITS).

Based on the data availability, the year of study is restricted for the period 1994–2015. Panel data was chosen to be used for the study because the data available was not considered long enough to be appropriate for time-series study. Panel study shall enable us to include more observations as well as raise the statistical strength and inference of the model (Baltagi, 2001).

Models Used

The various econometrics techniques applied to observe the nature of relationship between ET on RCA of food and food product, one of the top environmental sensitive products for G20 nations, are discussed below. Further, to account for the robustness of the model, panel ARDL test is applied to study the presence of long-run cointegration between the variables under study. Appropriate tests are also employed to check cross-sectional dependence and stationarity.

Trend and Correlation Analysis

Initially, trend analysis is done with the help of tabular and graphical representation to understand the basic nature of export percentage, RCA and ET of the G20 nations. Apart from trend, correlation coefficient has been estimated between RCA and ET to check the significant level of association between the variables by Pearson method (Equation 1).

r = \frac{\sum (x - \bar{x}) (y - \bar{y})}{\sqrt{\sum {(x - \bar{x})}^{2} \sum {(y - \bar{y})}^{2}}}

(1)

where r = r_xy = Pearson correlation coefficient.

Cross-sectional Dependence Test

The panel data often contains cross-sectional dependence among different cross-sectional units. So, it becomes necessary to ensure the presence of Cross-sectional Dependence Test (CSD) before applying unit root tests, Also, to identify the relevant models that can be applied for the dataset considered, it becomes crucial to check CSD so as to avoid any misspecifications in the model. In order to check CSD among the units, we have applied Breusch–Pagan LM (Breusch & Pagan, 1980), Pesaran LM and CD (Pesaran, 2004) tests, with the following hypothesis:

H₀: No cross-section dependence in residuals.

H₁: Cross-section dependence in residuals.

The Breusch and Pagan Lagrange Multiplier (LM) test is extensively used to examine the cross-sectional dependence in the models. The standard model is as follow:

Y_{i t} = α_{i} + β_{i} X_{i t} + μ_{i t}, (for i = 1_{\dots \dots} . N and t = 1, \dots, T)

(2)

where i and t are the cross-section and time dimensions respectively, while $\underline{α}$ and β indicate the individual intercepts and slope coefficients, respectively.

Breusch & Pagan’s LM Statistic is given by,

LM = T \sum_{i = 1}^{N - 1} \cdot \sum_{j = 1 + i}^{N} \hat{ρ} i j for i \neq j

(3)

where ${\hat{p}}^{2} j$ is the sample estimate of the pair-wise correlation of the residuals as follow,

{\hat{ρ}}_{i j} = {\hat{ρ}}_{j i} = \sum_{t = 1 \cdot}^{T} {\hat{u}}_{i t} {\hat{u}}_{j t}

(4)

{(\sum_{t = 1 \cdot}^{T} {\hat{u}}_{i t}^{2})}^{1 / 2} {(\sum_{t = 1 \cdot}^{T} {\hat{u}}_{j t}^{2})}^{1 / 2}

The scaled variant of LM test can be calculated as follow:

C D L M = (1 / N {(N - 1)}^{1 / 2} \sum_{i = 1 \cdot}^{N - 1} \sum_{k = i + 1 \cdot}^{N} (T {\hat{ρ}}^{2}_{i k} - 1)

(5)

A more general test that is said to be more reliable, proposed by Pesaran is as follow:

CD = \sqrt{2 T / N (N - 1)} (\sum_{i = 1 \cdot}^{N - 1} \sum_{j = i + 1 \cdot}^{N} \hat{ρ} i j)

(6)

Panel Unit Root

Panel unit root test is very common and is widely used in most empirical studies. Before estimating the long-run relationship between variables, it is necessary to determine the order of integration that is stationarity for each variable under study. In our analysis, different unit root tests given by Levin et al. (2002), Im et al. (2003) test and Fisher’s have been reviewed based on the following hypotheses:

Null hypothesis, H₀ = The data series is non-stationary and contains a unit root.

Alternative hypothesis, H₁ = The data series is stationary and doesn’t contain unit root.

Levin–Lin–Chu (LLC) considers the model as specified,

Δ y_{i t} = β_{i} + ρ_{j} y_{i, t - 1} + \sum_{z = 1}^{p i} δ_{i, z} Δ y_{i, t - z} + u_{i t}

(7)

where i=1, .., N and t = 1, .., T.

and error term ‘u_it’ is presumed to be independently distributed.

The Im–Pesaran–Shin (IPS) test is based on the assumptions of cross-sectional independence. The model has been constructed is as follows:

Δ y_{i t} = β_{i} + ρ_{j} y_{i, t - 1} + \sum_{z = 1}^{p i} δ_{i, z},Δ y_{i, t - z} + u_{i t}

(8)

The Fisher’s test of stationarity uses p-values for each cross-section (Choi, 2001). The test formula defined is as follows:

P_{MW} = - 2 \sum_{i = 1}^{N} \ln p_{i}

(9)

Panel ARDL Approach

Based on the order of integration of variables, panel ARDL model is used to study the short-run as well as the long-run effects of independent variable on the dependent variable. Panel ARDL model (Pesaran & Shin, 1999) can be applied whether the underlying variables are I(0) or I(1) or a mix of both. Although ARDL cointegration technique does not require to run stationarity tests but to avoid the failure of the model because of the presence of variable stationary at I(2), unit root test is carried out (Nkoro & Kelvin Uko, 2016).

RCA indicates whether a country is in the process of extending the products in which it has a trade potential. If RCA exceeds unity, the country is said to have a RCA in the product and vice versa. With this concept, the G20 nations whose RCA observed greater than one (RCA >1) in at least in a single year under study is considered for long-run cointegration analysis by ARDL model. By this process, 12 countries namely Australia, Canada, Brazil, France, South Africa, India, Germany, Mexico, Italy, Turkey, United States and the United Kingdom are considered for our panel ARDL analysis.

The ARDL model takes into account the problem of collinearity by considering the lag values of dependent variable along with independent variables and their lag values as well. It also tests for cointegration and forecasts long and short term relations while the variables are blend of I(0) and I(1). Let us assume that our data comprises of time periods t =1, ..., T and cross-section units i = 1, ..., N and panel ARDL (p, q) model is given as follows:

Y_{i, t} = \sum_{i = 1}^{p} α_{i j} Y_{i, t - j} + \sum_{i = 1}^{q} {β^{'}}_{i j} X_{i, t - j} + u_{i} + ε_{i t}

(10)

where X_i,t is the explanatory variable.

α_i,j, are the coefficients of the lagged variables.

Where the dependent variable is RCA and the lagged values of which are used as regressors, while X_i,t variables include ET. The selection of the model is based on lag length, selected by Akaike Information Criterion. The lag length is selected automatically by software itself. The pooled mean group (PMG) estimator provides the results for short run and long run, separately. The value of error correction term (ECT) can also be observed in the model that shows the convergence or divergence of the variables in estimating the long-run effects.

The model for (PMG estimator is given as

Δ y_{i, t} = θ_{i} + (E C_{i, t}) + \sum_{j = 1}^{p - 1} Ω_{i, j} Δ Y_{i, t - j} + \sum_{j = 0}^{q - 1} γ_{i, j} Δ X_{i, t - j} + ε_{i t}

(11)

where E C_{i, t} = y_{i, t - 1} - X_{j, t} β

(12)

The first part of the equation gives the ECT. The sign of ECT shows either convergence or divergence of the variables in the model. The negative value indicates a long-run relation among the dependent and explanatory variable in the model. The remaining part shows the lagged dependent variables used as regressors and lagged explanatory variables. The symbol ‘θ’ in the equation is termed as ECT while ‘β’ represents the coefficient in the long run.

IV. Empirical Results and Findings

Trend Analysis

Findings

From Table 2 and Figure 1, it can be clearly outlined that India, Saudi Arabia, Russia, Argentina and Mexico have experienced a continuous rise in their exports share of food and food products. In countries including Australia, France, Japan, EU, South Korea, Italy, UK and USA, the export share has reduced continuously over time. Further, analysing the data using trend line, it can be observed that Brazil, Indonesia, China, Turkey and South Africa show an increasing trend whereas, Germany and Canada exhibit a falling trend.

Figure 1.

Source: The authors.

Similarly, in Table 3 and Figure 2, we can see the trend of RCA of food and food products for the G20 countries from 1994 to 2015. It is noticeable that in countries—Canada, Japan, Indonesia, Argentina, Saudi Arabia, Germany, Italy, Mexico, USA and Brazil, the value of RCA increases continuously and South Africa, Russia, France also show an increasing trend in the value of their RCA index. Whereas, Turkey, China, South Korea, EU whose RCA has fallen continuously and Australia, UK, India follow a decreasing trend over the period.

Figure 2.

Source: The authors.

Table 3.

RCA of G20 Countries.

S. No.	Country	RCA (1994)	RCA (2000)	RCA (2015)
1.	Australia	2.002547	2.481950	1.643657
2.	Canada	0.694927	0.881317	1.281012
3.	France	1.577585	1.775569	1.708169
4.	Germany	0.591392	0.671059	0.765462
5.	Italy	0.752745	1.024773	1.205254
6.	Japan^a	0.064366	0.079512	0.114023
7.	South Korea^a	0.262542	0.240376	0.185499
8.	Mexico	0.650755	0.71192	0.857806
9.	Turkey	2.231443	1.997156	1.441106
10.	United Kingdom	0.917096	0.95706	0.909015
11.	United States	0.97553	1.001351	1.048677
12.	Brazil	3.238974	3.716019	4.319386
13.	China^a	0.969855	0.808719	0.354177
14.	India	1.367143	1.627583	1.19124
15.	Russia^a	0.191794 (1996)	0.17402	0.518116
16.	South Africa	0.85353	1.368261	1.323897
17.	Argentina^a	NA	6.188564	7.448431
18.	Indonesia^a	NA	1.144926	2.250121
19.	EU^a	NA	0.972295	0.937801
20.	Saudi Arabia^a	NA	0.078369	0.231142

Source: Authors’ calculations.

Note: NA: Not available.

^aRCA < 1.

Additionally, the trend of environmental tax can be seen in Table 4 and Figure 3. The value of tax in France, Japan, South Korea, China and Russia has shown a continuous increase with Australia, Canada, Italy, US showing increasing trend. The value of environmental tax, in Turkey and UK seems to be constant throughout the period. Countries including Germany, Mexico and South Africa have shown a continuous fall, with Brazil and India showing a falling trend.

Figure 3.

Source: The authors.

Table 4.

Environmental Tax of G20 Countries.

Country	1994	2000	2015
Australia	1.5	1.75	1.5
Canada	1.5	1.25	2.25
France	1.5	1.75	3
Germany	1.5	1.25	1
Italy	1.5	2.75	2.5
Japan^a	2.5	2.75	3
South Korea^a	0.75	0.75	2.5
Mexico	1	0.75	0.25
Turkey	1.25	1.25	1.25
United Kingdom	1.5	1.5	1.5
United States	0.75	1.75	1
Brazil	0.5	0.25	0.25
China^a	0.5	0.5	1.75
India	0.5	0.75	0.25
Indonesia^a	0	0	0
Russia^a	0.5	0.5	1
South Africa	1	0.75	0.5
EU^a	NA	NA	NA
Saudi Arabia^a	NA	NA	NA
Argentina^a	NA	NA	NA

Source: OECD database.

Note: NA: Not available.

^aRCA < 1.

The results based on Trend Analysis show that the comparative advantage of food manufacturing industries depends on the stringency level in respective country. Mainly the countries (Australia, France, Canada, Japan, South Korea, USA and Italy) with falling exports and strict environmental policies shift to the countries with less stern policies, thus making them pollution havens. Further, the countries (Turkey and UK) with falling comparative advantage regardless of stringency could be due to various other factors playing an important role in influencing the comparative advantage of the country in food industries. Such other factors may include availability of cheap factors of production, increased efficiency leading to specialisation and economies of scale. China and South Korea are the countries where negative impact of tax is clearly visible. There are countries where stringency plays a significant role in improving the comparative advantage of food industries (Germany, Mexico, Brazil and South Africa). Some countries among the selected G20 are facing stringency and simultaneously enjoying comparative advantage as well, indicating that such countries have effectively adopted PH and have therefore attained a balance between trade and environmental quality (France, Russia, Japan, Canada, USA and Italy). Else, it can be concluded that the environmental policies in respect of food manufacturing do not play much significant role in some countries since the environmental damages due to food-processing activities are not considered of much importance and are therefore neglected as food industry plays a highly significant role in everyday life. However, Australia is the only country where increase in environmental tax causes RCA of food and food products to fall and was recently labelled as one with great capacity to reduce food-related GHG emissions. Meat and dairy production are two primary sources, according to the Intergovernmental Panel on Climate Change (Parletta, 2019). So far as India is concerned, it can be seen from the results that irrespective of falling stringency, the RCA is still falling, though exports are rising. It can only be concluded that India, with rising inflation, lack of efficient techniques and having no strict laws regarding food manufacturing, is moving towards the path of becoming a pollution haven.

Correlation Analysis

Findings

Table 5 reports the country wise correlation coefficients. It is observed that India, Australia, Canada, Italy, Turkey, Russia and France show positive correlation. Out of these countries, Canada has the most significant and strong positive correlation (0.77) between the variables. On the other hand, countries including Brazil, Mexico, US, UK, South Korea, Japan, China, Germany and South Africa exhibit negative correlation, where China (−0.88) has strong negative correlation between RCA and ET.

Table 5.

Correlation Table.

Country	Correlation Coefficient	Direction of Correlation
Brazil	–0.17	Negative
India	0.20	Positive
Australia	0.53	Positive
Italy	0.67	Positive
Turkey	0.19	Positive
Mexico	–0.78	Negative
UK	–0.10	Negative
USA	–0.61	Negative
Germany	–0.79	Negative
South Africa	–0.06	Negative
France	0.16	Positive
Canada	0.77	Positive
South Korea^a	–0.76	Negative
Russia^a	0.48	Positive
China^a	–0.88	Negative
Japan^a	–0.19	Negative

Source: Authors’ calculations.

Note: ^aRCA < 1.

From the results of the Correlation Analysis, it can be seen that out of countries exhibiting positive correlation, Canada has the most significant and strong positive correlation between the variables. Therefore, we can conclude that environmental policies in a developed country like Canada are strictly implemented and sincerely followed by the food manufacturing units. Thereby, abiding to porter principles. Also it can be estimated from the results that the Canadian government takes the environmental concerns and emission generating from food industries seriously, ensuring proper rules are being implemented and followed up.

On the other hand, countries exhibiting negative correlation, where China has strong negative correlation between RCA and ET, implying that falling tax causing RCA to rise, concluding that in these countries, food related emissions are not considered to be of much relevance from environmental point of view.

Cross-section Dependence

Findings

The estimated result for cross-section dependence test for panel data is presented in Table 6. Since we have large time series and small cross-sections, the LM test become the most valid and reliable. As shown in Table 6, the BP-LM test statistic and Pearson scaled LM test statistic is 312.7206 and 21.47427, respectively with a p-value of .0000 for both. At 1% level of significance, we reject the H₀ and conclude that there is no CSD in the data. It means that a country’s comparative advantage and environmental tax (stringency) do not influence the other country’s RCA and ET. Each G20 nation implements its own policies without influencing the regulations of the other G20 countries.

Table 6.

Cross-sectional Dependence Test.

Test	Statistic	Prob.
Breusch–Pagan LM Pesaran scaled LM Pesaran CD	312.7206 21.47427 2.863978	0.0000* 0.0000* 0.0042*

Source: Authors’ calculations.

Notes: Null hypothesis: No cross-section dependence in residuals.

* shows significance at 1% level.

Panel Unit Root Test Results

Findings

Before evaluating the relation between RCA and the environmental tax, we need to first assess the stationarity of the variables, otherwise our model will give spurious results. In order to check stationarity, we apply four different kinds of unit root tests, LLC, IPS, Fisher–ADF and Fisher–PP tests. Among the tests, LLC given by Levin et al. (2002) is assumed to be as common unit root test. The IPS, proposed by Im et al. (2003) and Fisher’s test are carried out as individual unit root tests. The null hypothesis used in all the tests is the same as that is the existence of unit root.

Table 7 gives the unit root test results for the variables. For RCA, the statistically significant results at level confirm that the panel is stationary. However, at the first difference, significant results confirm that ET is stationary. Therefore, our test result confirms that both variables are stationary at different order. Thus, we can use panel ARLD approach.

Table 7.

Panel Unit Root Tests.

Dependent Variable Method	RCA: I(0)		TAX: I(1)
Dependent Variable Method	Statistic	Prob.	Statistic	Prob.
Null: Unit root (assumes common unit root process) Levin–Lin–Chu test	–3.72306	0.0001	–11.6904	0.0000*
Null: Unit root (assumes individual unit root process) Im–Pesaran–Shin W-stat ADF–Fisher chi-square PP–Fisher chi-square	–1.95826 39.7350 46.8837	0.0251 0.0228 0.0035	–11.2513 152.056 194.493	0.0000* 0.0000* 0.0000*

Source: Authors’ calculations.

Notes: Automatic lag length selection based on SIC.

* shows significance at 1% level.

If the p-value comes to be less than .05, it can be said that the data are stationary at 5% significance. The unit root test results show that we can reject the H₀ for both variables, since p-values are less than 5%. Accepting the H₀ means that they do not have unit root and the results of all the methods employed significantly indicate that both the series are stationary but at different orders.

Panel ARDL

Findings

Table 8 summarises the results for Panel ARDL test. The findings of the study show that RCA, being our dependent variable is significant in the long run, hence giving indication to reject the null hypothesis and accept the alternate hypothesis concluding that environmental tax happens to influence RCA of food and food products, at 5% level of significance. It can be seen that, in the long run, the coefficient of tax is negative (−0.0667) implying that a unit increase in tax, decreases the RCA by 0.0667 units.

Table 8.

ARDL Results (Selected Model: ARDL (1, 1)).

Variable	Coefficient	Std Error	t-Statistic	Prob.
Long run
TAX	–0.066735	0.030479	–2.189535	0.0296**
Short run
COINTEQ01	–0.293002	0.075039	–3.904693	0.0001*
D(TAX)	–0.039967	0.039465	–1.012733	0.3123
C	0.444916	0.122738	3.624914	0.0004*

Source: Authors’ calculations.

Note: *,** shows significance at 1% and 5%, respectively.

For short-run analysis, Table 6 reports the Cointegrating Equation which is negative (−0.2930) and significant (Probability = 0.0001). This result implies that the variables converge with a speed of convergence being 29.30%, towards the long-run equilibrium. The p-value of our variable (TAX) is not significant, implying that, in the short period, the environmental tax does not have any significant impact on RCA. Also, we find that the coefficient of the intercept is positive (0.4449) and significant (p-value < .05) meaning that, in the short run, without taking the impact of environmental tax, the RCA of G20 countries, amounts to 0.4449.

Through the results of Panel ARDL, it can be interpreted that the G20 exports of food and food products for selected countries are highly dependent on the level of stringency, in the long run. The selected countries may follow strict regulations to save the environment from further degradation caused by toxic and lethal chemicals and effluents released from the food manufacturing industries which happen to negatively impact the environment that is more clearly realised in the longer period.

So far as short period is concerned, the stringent policies do not play any vital role in influencing the export competitiveness of the respective G20 country since the industries adapt to regulation gradually.

Cross-section Short-run Effects

Findings

Table 9 reports the short-run effect for each cross-sectional unit. It can be observed that there is a movement towards the long run in all the countries since coefficient of the cointegration equation is negative and significant for all. The findings reveal that, only in India and Germany, the short-run impact of environmental tax on RCA is insignificant at 5% level of significance (p-values > .05). It is also noted that a unit increase in tax improves the RCA by 0.0630 units in Brazil, 0.0338 units in Canada, 0.0008 in Germany, 0.0275 in UK, 0.0476 in Italy and 0.0883 in Turkey. On the other hand, a unit increase in tax reduces the RCA by 0.4263 in Australia, 0.048261 in India, 0.0874 in France, 0.0213 in Mexico, −0.0329 in USA, and −0.1245 in South Africa.

Table 9.

Cross Section Short Run Effects.

Country	Variable	Coefficient	Standard Error	t-statistic	Probability
Brazil	COINTEQ01 D(TAX) C	–0.354412 0.063062 1.485850	0.011287 0.009131 0.185665	–31.40041 6.906234 8.002869	0.0001* 0.0062* 0.0041*
Canada	COINTEQ01 D(TAX) C	–0.038215 0.033800 0.068973	0.001689 0.001088 0.002109	–22.63120 31.07003 32.71097	0.0002* 0.0001* 0.0001*
Australia	COINTEQ01 D(TAX) C	–0.101601 –0.426324 0.187394	0.009141 0.045523 0.038918	–11.11547 –9.365097 4.815066	0.0016* 0.0026* 0.0171*
India	COINTEQ01 D(TAX) C	–0.190118 –0.048261 0.264546	0.021480 0.036719 0.045999	–8.850873 –1.314322 5.751125	0.0030* 0.2802 0.0104*
France	COINTEQ01 D(TAX) C	–0.241704 –0.087476 0.462189	0.014067 0.001583 0.047931	–17.18194 –55.24263 9.642829	0.0004* 0.0000* 0.0024*
Germany	COINTEQ01 D(TAX) C	–0.148061 0.000811 0.125789	0.008630 0.001823 0.005624	–17.15613 0.444652 22.36575	0.0004* 0.6867 0.0002*
Mexico	COINTEQ01 D(TAX) C	–0.323082 –0.021305 0.267956	0.025915 0.001388 0.017075	–12.46686 –15.35002 15.69284	0.0011* 0.0006* 0.0006*
USA	COINTEQ01 D(TAX) C	–0.499003 –0.032956 0.570868	0.040384 0.000758 0.056549	–12.35645 –43.45267 10.09512	0.0011* 0.0000* 0.0021*
UK	COINTEQ01 D(TAX) C	–0.972340 0.027530 1.001411	0.052675 0.004413 0.056815	–18.45934 6.238376 17.62575	0.0003* 0.0083* 0.0004*
South Africa	COINTEQ01 D(TAX) C	–0.444865 –0.124560 0.605667	0.021663 0.028873 0.038692	–20.53581 –4.313986 15.65353	0.0003* 0.0229** 0.0006*
Italy	COINTEQ01 D(TAX) C	–0.067227 0.047693 0.098936	0.001863 0.000719 0.002677	–36.08254 66.36045 36.95883	0.0000* 0.0000* 0.0000*
Turkey	COINTEQ01 D(TAX) C	–0.135400 0.088382 0.199418	0.006368 0.007569 0.020345	–21.26108 11.67716 9.801887	0.0002* 0.0013* 0.0023*

Source: Authors’ calculations.

Note: *,** shows significance at 1% and 5%, respectively.

From the results of short-run cross-section effects, it can be said that, in India and Germany, the environmental policies are not much stringent to strongly impact the comparative advantage of food manufacturing units in these nations, mainly because of absence of direct laws targeting food industries. It is also noted that a unit increase in stringency improves the RCA in Brazil, Canada, Germany, UK, Italy and Turkey, implying that these G20 countries are moving towards the following PH or in the absence of direct food laws, there may be other factors like availability of less expensive factors of production due to prevalence of low market wage rate, so far as short run is considered. On the other hand, in short run, a unit increase in tax reduces the RCA in Australia, India, France, Mexico, USA and South Africa, indicating that these G20 economies have started implementing such policies that affect their food industry’s competitiveness in the international markets that may be effective in the short run but may lack long-run effectiveness due to lack of government incentives and resources.

V. Conclusions and Policy Recommendations

Conclusion

The main objective of our study was to understand the relationship between RCA on food and food products industry and environmental tax in the G20 countries. Trend analysis, correlation, panel ARDL and short-run cross-section effects analysis are conducted for the present study. Out of 20 countries, 12 countries have been considered for ARDL analysis as the RCA was below one for rest of the countries.

It can be concluded from the study that rising and falling of exports share mainly depend on stringency levels that vary across countries. The developing countries can be regarded as pollution havens with lower levels of stringency show rising trend in exports share whereas, the developed countries may face tough environmental policies exhibit a falling trend, as evident from the results.

The countries with increasing RCA enjoy comparative advantage, irrespective of environmental tax imposed, which could be due to availability of cheap factors of production, increased efficiency leading to specialisation and economies of scale.

Some countries among the selected G20 are facing stringency and simultaneously enjoying comparative advantage as well, indicating that such countries have effectively adopted PH and have therefore attained a balance between trade and environmental quality. Else, it can be concluded that the environmental policies in respect of food manufacturing do not play much significant role in some countries since the environmental damage due to food processing activities are not considered of much importance and are therefore neglected as food industry plays a highly significant role in everyday life.

In the long run, environmental stringency happens to influence RCA of food and food products. Whereas, in the short run, stringency does not affect RCA to any great extent in G20 countries.

Recommendations

This study recommends that appropriate and sound policies regarding proper resource allocations and trade policies that promote exportation of goods and international trade in order to bring positive impacts on G20 economies in both short run and long run, simultaneously keeping in balance with environmental quality. The examination of the trade and environment relationship reveals that the responsibility of the G20 countries should be to combat climate change by reducing the emissions from food industries through the international trade. The G20 being a major player in the global food emissions, it is suggested that countries facing less stringent or no direct taxes related to pollution caused by food industries, a specific tax should be imposed on G20 food industries crossing a particular emission limit set by the authorities.

Also, in countries which are facing stringent environmental laws and taxes related to food production, tax incentives should be given to food manufactures to stop tax evasion, if any, and encourage them to survive more efficiently in the competitive market and focus on improving their comparative advantage. At the same time, by providing tax credits to G20 food industries, it will boost them to serve their own country and lead it to progress regardless of relocating due to high tax burden.

Footnotes

Authors Contributions

All authors contributed to the study. Material preparation, data collection and analysis were performed by Alisha Mahajan. The first draft of the manuscript was written by Alisha Mahajan and the second author Dr Kakali Majumdar commented on the work done. Both the authors read and approved the final manuscript.

Declaration of Conflicting Interests

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Alisha Mahajan

References

Arnold

F. S.

(1999). Environmental protection: Is it bad for the economy? U.S. Environmental Protection Agency.

Baltagi

B. H.

(2001). Econometric analysis of panel data (2nd ed.). John Wiley & Sons.

Bank

World Ecofys Econom.

Vivid (2016). State and trends of carbon pricing . World Bank. 10.1596/978-1-4648-1001-5

Baumol

W. J.

Oates

W. E.

(1988). The theory of environmental policy . Cambridge University Press.

Boden

Marland

Andres

(2011). Global, regional, and national fossil-fuel co2 emissions, 1751–2008 (Version 2011). Environmental System Science Data Infrastructure for a Virtual Ecosystem; Carbon Dioxide Information Analysis Center (CDIAC), Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States). https://doi.org/10.3334/CDIAC/00001_V2011

Bonnet

Bouamra-Mechemache

Corre

(2018). An environmental tax towards more sustainable food: Empirical evidence of the consumption of animal products in France. Ecological Economics , 147, 48–61. https://doi.org/10.1016/j.ecolecon.2017.12.032

Bordigoni

Hita

Le Blanc

(2012). Role of embodied energy in the European manufacturing industry: Application to short-term impacts of a carbon tax. Energy Policy , 43, 335–350. https://doi.org/10.1016/j.enpol.2012.01.011

Breusch

T. S.

Pagan

A. R.

(1980). Lagrange multiplier test and to model applications specification in econometrics. Review of Economic Studies , XLVII(1), 239–253.

Cagatay

Mihci

(2006). Degree of environmental stringency and the impact on trade patterns. Journal of Economic Studies , 33(1), 30–51. https://doi.org/10.1108/01443580610639884

10.

Cheng

Dai

Wang

Zhao

Masui

(2015). Impacts of carbon trading scheme on air pollutant emissions in Guangdong Province of China. Energy for Sustainable Development , 27, 174–185. https://doi.org/10.1016/j.esd.2015.06.001

11.

Choi

(2001). Unit root tests for panel data. Journal of International Money and Finance , 20(2), 249–272. https://doi.org/10.1016/S0261-5606(00)00048-6

12.

Clune

Crossin

Verghese

(2017). Systematic review of greenhouse gas emissions for different fresh food categories. Journal of Cleaner Production , 140, 766–783. https://doi.org/10.1016/j.jclepro.2016.04.082

13.

Copeland

B. R.

Taylor

M. S.

(2004). Trade, growth, and the environment. Journal of Economic Literature , 42(1), 7–71.

14.

Costantini

Mazzanti

(2012). On the green and innovative side of trade competitiveness? The impact of environmental policies and innovation on EU exports. Research Policy , 41(1), 132–153. https://doi.org/10.1016/j.respol.2011.08.004

15.

Dmitriev

O. N.

Novikov

S. V.

(2018). Transformation of russian tax system as part of the integration of the economy into international high-technology production field of socio-economic systems. European Research Studies Journal , XXI(Issue 4), 400–413. https://doi.org/10.35808/ersj/1130

16.

Ederington

(2007). Nafta and the pollution haven hypothesis. Policy Studies Journal , 35(2), 239–244. https://doi.org/10.1111/j.1541-0072.2007.00218.x

17.

Gao

Yao

Wang

Liu

(2019). Dynamic effect of environmental tax on export trade: Based on DSGE mode. Energy & Environment , 30(7), 1275–1290. https://doi.org/10.1177/0958305X19842380

18.

Zou

Qiao

Chen

Wang

Luo

Ning

(2019). Does the double dividend of environmental tax really play a role in OECD countries? A study based on the panel ARDL model. Ekoloji , 28(107), 49–62.

19.

K. S.

Pesaran

M. H.

Shin

(2003). Testing for unit roots in heterogeneous panels. Journal of Econometrics , 115(1), 53–74. https://doi.org/10.1016/S0304-4076(03)00092-7

20.

Khan

Sisi

Siqun

(2019). Environmental regulations an option: Asymmetry effect of environmental regulations on carbon emissions using non-linear ARDL. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects , 41(2), 137–155. https://doi.org/10.1080/15567036.2018.1504145

21.

Kozluk

Timiliotis

(2016). Do environmental policies affect global value chains? A new perspective on the pollution haven hypothesis (OECD Economics Department Working Papers, No. 1282). OECD.

22.

Levin

Lin

C.-F.

Chu

James C.-S.

(2002). Unit root tests in panel data: Asymptotic and finite-sample properties. Journal of Econometrics , 108(1), 1–24. https://doi.org/10.1016/S0304-4076(01)00098-7

23.

J. F.

Lin

Z. X.

(2015). The impact of sales tax on economic growth in the United States: An ARDL bounds testing approach. Applied Economics Letters , 22(15), 1262–1266. https://doi.org/10.1080/13504851.2015.1023933

24.

Chen

Zeng

Shao

Xie

Chen

Cao

(2014). Biomass burning contribution to ambient volatile organic compounds (Vocs) in the Chengdu–Chongqing Region (Ccr), China. Atmospheric Environment , 99, 403–410. https://doi.org/10.1016/j.atmosenv.2014.09.067

25.

Mardones

Baeza

(2018). Economic and environmental effects of a CO2 tax in Latin American countries. Energy Policy , 114, 262–273. https://doi.org/10.1016/j.enpol.2017.12.001

26.

Matsumoto

Masui

(2011). Analyzing long-term impacts of carbon tax based on the imputed price, applying the AIM/CGE model. Management of Environmental Quality: An International Journal , 22(1), 33–47. https://doi.org/10.1108/14777831111098462

27.

Mittal

Sharma

2003. Anthropogenic Emissions from Energy Activities in India: Generation and Source Characterization, Part I. www.osc.edu/research/archieve/pcrm/emissions

28.

Moberg

Andersson

Walker

Säll

Hansson

S. P.-A.

Röös

(2019). Determining the climate impact of food for use in a climate tax –Design of a consistent and transparent model. The International Journal of Life Cycle Assessment , 24(9), 1715–1728. https://doi.org/10.1007/s11367-019-01597-8

29.

Morgenstern

R. D.

Shih

J.-S.

Zhang

(2004). The near-term impacts of carbon mitigation policies on manufacturing industries. Energy Policy , 32(16), 1825–1841. https://doi.org/10.1016/S0301-4215(03)00171-X

30.

Muradian

Martinez-Alier

(2001). Trade and the environment: From a ‘Southern’ perspective. Ecological Economics , 36(2), 281–297. https://doi.org/10.1016/S0921-8009(00)00229-9

31.

Nkoro

Uko

Kelvin A.

(2016). Autoregressive distributed lag (ARDL) cointegration technique: Application and interpretation. Journal of Statistical and Econometric Methods , 5(4), 63–91.

32.

Ntombela

S. M.

Bohlmann

H. R.

Kalaba

M. W.

(2019). Greening the South Africa’s economy could benefit the food sector: Evidence from a carbon tax policy assessment. Environmental and Resource Economics , 74(2), 891–910. https://doi.org/10.1007/s10640-019-00352-9

33.

Parletta

Natalie.

(2019). What would a ‘climate diet’ look like in Australia? https://www.theguardian.com/food/2019/oct/29/what-would-a-climate-diet-look-like-in-australia

34.

Pesaran

M. H.

(2004). General diagnostic tests for cross section dependence in panels (Working Papers in Economics No. 0435). University of Cambridge, Faculty of Economics.

35.

Pesaran

Shin

(1999). An Autoregressive Lag Modeling Approach to Cointegration Analysis. In Strom

. (ed) Econometrics and Economic Theory in 20th Century: The Ragnar Frisch centennial Symposium , Cambridge University Press, Cambridge.

36.

Piwowar

(2019). Low-carbon agriculture in Poland: Theoretical and practical challenges. Polish Journal of Environmental Studies , 28(4), 2785–2792. https://doi.org/10.15244/pjoes/92211

37.

Poore

Nemecek

(2018). Reducing food’s environmental impacts through producers and consumers. Science , 360(6392), 987–992.

38.

Porter

M. E.

(1991). Towards a dynamic theory of strategy. Strategic Management Journal , 12(S2), 95–117. https://doi.org/10.1002/smj.4250121008

39.

Porter

M. E.

Linde

C. van der.

(1995). Toward a new conception of the environment-competitiveness relationship. Journal of Economic Perspectives , 9(4), 97–118. https://doi.org/10.1257/jep.9.4.97

40.

Rivers

Schaufele

(2015). The effect of carbon taxes on agricultural trade: Effect of carbon taxes on agricultural trade. Canadian Journal of Agricultural Economics/Revue Canadienne d’agroeconomie , 63(2), 235–257. https://doi.org/10.1111/cjag.12048

41.

Robson

(2014). Australia’s carbon tax: An economic evaluation: Australia’s carbon tax: An economic evaluation. Economic Affairs , 34(1), 35–45. https://doi.org/10.1111/ecaf.12061

42.

Säll

Gren

I.-M.

(2015). Effects of an environmental tax on meat and dairy consumption in Sweden. Food Policy , 55, 41–53. https://doi.org/10.1016/j.foodpol.2015.05.008

43.

Sebastian. (2015, November 6). G20 emissions. Climate Transparency . https://www.climate-transparency.org/g20-emissions

44.

Sedehi

Esfahanian

(2019). The effect of green taxes on labor productivity in the Iranian economy. Iran Economic Review , 23(4), 1041–1055.

45.

Silajdzic

Mehic

(2017). The impact of environmental taxes on competitive performance of pollution-intensive industries among transition economies: Evidence from panel analysis. In Bilgin

M. H.

Danis

Demir

Can

(Eds), Financial environment and business development (Vol. 4, pp. 155–169). Springer International Publishing. https://doi.org/10.1007/978-3-319-39919-5_13

46.

Springmann

Mason-D’Croz

Robinson

Wiebe

Godfray

H. C. J.

Rayner

Scarborough

(2017). Mitigation potential and global health impacts from emissions pricing of food commodities. Nature Climate Change , 7(1), 69–74. https://doi.org/10.1038/nclimate3155

47.

Tewari

Pillai

(2005). Global standards and the dynamics of environmental compliance in India’s leather industry. Oxford Development Studies , 33(2), 245–267. https://doi.org/10.1080/13600810500137947

48.

Tian

Dai

Geng

Huang

Masui

Fujita

(2017). The effects of carbon reduction on sectoral competitiveness in China: A case of Shanghai. Applied Energy , 197, 270–278. https://doi.org/10.1016/j.apenergy.2017.04.026

49.

US EPA, O. (2015, July 16). Toxics release inventory (Tri) national analysis [Collections and Lists] . https://www.epa.gov/trinationalanalysis

50.

Zhao

Y.-H.

(2011). The study of effect of carbon tax on the international competitiveness of energy-intensive industries: An empirical analysis of OECD 21 countries, 1992–2008. Energy Procedia , 5, 1291–1302. https://doi.org/10.1016/j.egypro.2011.03.225

Impact of Environmental Tax on Comparative Advantage of Food and Food Products: A Study of G20 Countries in Light of Environmentally Sensitive Goods

Abstract

Keywords

1. Introduction

Trade and Environment Debate

Food and Products—An Environmentally Sensitive Industry

Table 1.

Intensity and Total CO2 Emissions Embodied in Gross Exports for G20.

Environmental Stringency and Environmental Tax

Environmental Taxes in G20

Models and Methods of Measuring Impacts of Environmental Tax

General Equilibrium (GE) Model

Partial Equilibrium Model

Input–Output Model

Regression Analysis

Gravity Model

Autoregressive Distributed Lag Technique

III. Data and Methodology

Description About Sample and Variables Under Study

Table 2.

Food and Food Products Export: G20 Countries.

Models Used

Trend and Correlation Analysis

Cross-sectional Dependence Test

Panel Unit Root

Panel ARDL Approach

Trend Analysis

Findings

RCA of G20 Countries.

Environmental Tax of G20 Countries.

Correlation Analysis

Findings

Correlation Table.

Cross-section Dependence

Findings

Cross-sectional Dependence Test.

Panel Unit Root Test Results

Findings

Panel Unit Root Tests.

Panel ARDL

Findings

ARDL Results (Selected Model: ARDL (1, 1)).

Cross-section Short-run Effects

Findings

Cross Section Short Run Effects.

Conclusion

Recommendations

Footnotes

Authors Contributions

Declaration of Conflicting Interests

Funding

ORCID iD

References

Intensity and Total CO₂ Emissions Embodied in Gross Exports for G20.