Elsevier

Economic Modelling

Volume 100, July 2021, 105498
Economic Modelling

Convergence in OPEC carbon dioxide emissions: Evidence from new panel stationarity tests with factors and breaks

https://doi.org/10.1016/j.econmod.2021.105498Get rights and content

Highlights

  • Convergence in CO2 emissions across countries affects climate agreement negotiations.

  • We test stochastic convergence in CO2 emissions using new panel stationarity tests.

  • New panel stationarity tests provide less evidence of convergence in CO2 emissions.

  • Absence of convergence in CO2 emissions among OPEC countries requires policy actions.

  • Policy actions to reduce CO2 emissions are discussed.

Abstract

Convergence in per capita emissions occurs when countries with higher initial levels of per capita emissions experience lower emission growth than countries with lower initial levels of per capita emissions. The recent literature adopts the notion of stochastic convergence, which implies that per capita emissions tend to converge to a stationary level over time. The empirical literature on emissions convergence has yielded mixed results. We present new panel stationarity tests that account for cross-correlations and structural breaks. These new tests are more general than previous tests for stochastic convergence. Using annual data for 13 OPEC countries and 18 non-OPEC oil-producing countries from 1960 to 2016, we find less evidence of stochastic convergence than previous studies. The absence of convergence in per capita emissions across countries could hinder the acceptance of climate agreements based on the allocation of per capita emissions obligations. Policy actions to address emissions growth are also discussed.

Introduction

The growing concerns regarding greenhouse gas emissions, of which carbon dioxide (CO2) emissions are a major component, have spawned a great deal of interest from both policymakers and researchers to understand how best to reduce emissions. As CO2 emissions are directly related to a number of factors that include a country’s energy mix, natural resource endowments, economic structure, level of economic development, and environmental policies, the issue of emissions reduction can become a contentious and political issue, as reflected by the Framework Convention on Climate Change in 1992, the Kyoto Protocol in 1997, and the Paris Agreement in 2015. However, a key component in the discussion of emissions mitigation strategies and the allocation of emissions obligations across countries depends, in part, on the convergence behavior of per capita CO2 emissions.

Convergence in per capita CO2 emissions occurs when countries with higher initial levels of per capita CO2 emissions exhibit a lower growth rate in emissions than countries with lower initial levels of per capita CO2 emissions in moving toward a steady-state. Thus, convergence in per capita CO2 emissions suggests that countries are more likely to endorse emission allocation strategies as the relocation of emissions-intensive industries or the transfer of resources through the international trading of carbon allowances may be less of an issue. However, as noted by Aldy (2006), the political aspects of emissions allocation strategies are relevant in the negotiations of multilateral climate change agreements. For example, developing countries with lower per capita CO2 emissions may demand that industrialized countries play a greater role in mitigation efforts when they tend to have higher per capita CO2 emissions.2

As discussed in the surveys by Petersson et al. (2014), Acar et al. (2018), and Payne (2020), a vast majority of the large multi-country studies that encompass countries that vary in both size and the level of economic development have yielded mixed results as to the presence of per capita CO2 emissions.3 On the other hand, studies that focus exclusively on the convergence of per capita CO2 emissions within particular country groupings, defined by either income classification, geographic region, or institutional structure (i.e., OECD, EU, or G20) are more likely to identify convergence.4,5

Given that OPEC controls 72 percent of the worlds’ proven crude oil reserves and 39 percent of the world’s crude oil production, it is surprising only two studies have investigated the convergence of per capita CO2 emissions for OPEC countries.6 Based on data from 1960 to 2003, Panopoulou and Pantelidis (2009) utilize club convergence techniques to show the absence of convergence in per capita CO2 emissions among OPEC countries. Using a non-parametric distribution approach, Ordas Criado and Grether (2011) also fail to identify convergence in per capita CO2 emissions among OPEC countries. Indeed, OPEC offers an interesting case study with respect to CO2 emissions convergence on several fronts. First, the end-use energy prices are heavily subsidized in the vast majority of OPEC member countries. As a result, the relatively lower prices serve as a disincentive for end-users to focus on energy conservation and the utilization of more energy-efficient technologies to curb emissions (Solarin and Lean, 2018). Second, in general, OPEC member countries are heavily reliant on fossil fuels as a percentage of their overall energy consumption and fossil fuel exports which constitute a relatively large percentage of total exports. Indeed, as Tajudeen and Wossnik (2020) note, the increase in energy intensity is linked to both energy inefficiency and shifts to energy-intensive activities. Hence, the challenge for policymakers is balancing the need to diversify the energy mix against an economic structure dependent on fossil fuels. Third, OPEC member countries also vary in their economic development level, as reflected by the World Bank income classification.7,8

The recent literature has adopted the notion of stochastic convergence, which implies that relative per capita CO2 emissions tend to converge to a steady-state or stationary level over time. As such, a battery of different stationarity tests have been considered in the literature. However, our study is the first to investigate the stochastic convergence of per capita CO2 emissions in OPEC countries through the use of new panel stationarity tests that control for possible cross-correlations using a factor model with structural breaks. Such analysis would be particularly meaningful in light of OPEC’s prominence in world energy markets alongside the limited number of studies on the convergence of CO2 emissions in OPEC member countries. In addition, we also examine 18 non-OPEC oil-producing countries as a comparison group along with a subset of OPEC countries defined by geographic proximity.9 Second, we contribute to the literature on testing the null hypothesis of stationarity in panel data by proposing new second-generation panel stationarity tests based on the combination of p-values. The new tests allow for structural breaks and cross-sectional dependence, and can be viewed as significant complements to the existing panel stationarity tests. As noted by Desli and Gkoulgkoutsika (2020) of the importance of evaluating the underlying deterministic trend, our consideration of structural breaks is a relevant consideration. Furthermore, our simulations indicate the new combination tests have better size properties than the group-mean tests and show similar power performance. We also develop the response surface function to obtain the corresponding p-value of the test statistic through a convenient and simple procedure.10 Therefore, one can easily draw inferences for each cross-sectional unit using the p-values from the response surface function.11

Our test results indicate that all the panel stationarity tests do not support the stochastic convergence of relative per capita CO2 emissions. The univariate test results of individual countries differ across cross-sections and appear to be mixed. The first-generation approach rejects the null hypothesis of stationarity more often than the second-generation approach for OPEC and non-OPEC oil-producing countries. Accounting for structural changes as a sharp process further results in more evidence on the absence of stochastic convergence for relative per capita CO2 emissions in non-OPEC oil-producing countries than in the case of OPEC and neighboring OPEC countries. When we allow for smooth/gradual structural breaks, the number of rejections of the null hypothesis of stationarity is dramatically increased from three to ten for OPEC countries. We also observe more rejections of the null hypothesis (from one to three) for neighboring OPEC countries. Nonetheless, in non-OPEC oil-producing countries, there is no considerable change in the results by moving from a sharp break approach to a smooth break approach.

The rest of the study is organized as follows. Section 2 describes the data for OPEC and non-OPEC oil-producing countries. Section 3 discusses the panel stationarity tests. Section 4 presents the empirical results. Section 5 provides concluding remarks.

Section snippets

Data

The Organization of Petroleum Exporting Countries (OPEC) has evolved over time in terms of member countries. Founded in 1960, the initial members of OPEC included Iran, Iraq, Kuwait, Saudi Arabia, and Venezuela. Subsequently, Qatar (1961), Indonesia (1962), Libya (1962), United Arab Emirates (1967), Algeria (1969), Nigeria (1971), Ecuador (1973), Gabon (1975), Angola (2007), Equatorial Guinea (2017), and the Congo (2018) joined as member countries. However, OPEC country membership has also

Panel stationarity tests

Paralleling the work of Carlino and Mills (1993) and Bernard and Durlauf (1995; 1996) in their examination of income convergence, we follow the same strategy. Formally, we can say that stochastic convergence in a country’s per capita CO2 emissions is present if relative per capita CO2 emissions are trend-stationary via unit root or stationarity tests. Here, relative emissions are defined as the log of per capita CO2 emissions for country i relative to the average of all countries in a

Empirical results and discussion

Before we proceed with the estimation results, it is important to clarify two issues. The first issue is that disturbances may have serial dependence and/or heteroscedasticity in practice. In order to allow for a general form of temporal dependence and heteroscedasticity, one can obtain individual tests (LMi) by using the consistent long-run variance ωˆεi2 based on either a parametric or a non-parametric estimator. Carrion-i-Silvestre and Sansó (2006) compare the small sample properties of

Concluding remarks

Despite the economic and geopolitical importance of OPEC, there have been a limited number of studies pertaining to the convergence of per capita CO2 emissions in OPEC countries. In this regard, this study examines the stochastic convergence of relative per capita CO2 emissions of the current 13 OPEC member countries. We also examine the stochastic convergence of relative per capita CO2 emissions for 18 non-OPEC oil-producing countries as a comparison group along with neighboring OPEC countries

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (67)

  • R. Ezcurra

    Is there cross-country convergence in carbon dioxide emissions?

    Energy Pol.

    (2007)
  • P.G. Fredriksson et al.

    Chasing the smokestack: strategic policymaking with multiple instruments

    Reg. Sci. Urban Econ.

    (2004)
  • M.J. Herrerias

    The environmental convergence hypothesis: Carbon dioxide emissions according to the source of energy

    Energy Pol.

    (2013)
  • S. Honma et al.

    An empirical investigation of the balance of embodied emission in trade: Industry structure and emission abatement

    Econ. Modell.

    (2020)
  • T. Jobert et al.

    Convergence of per capita dioxide emissions in the EU: Legend or reality?

    Energy Econ.

    (2010)
  • E. Karakaya et al.

    Replication of Strazicich and list (2003): Are CO2 emission levels converging among industrial countries?

    Energy Econ.

    (2019)
  • D. Kwiatkowski et al.

    Testing the null hypothesis of stationary against the alternative of a unit root

    J. Econom.

    (1992)
  • J. Lee et al.

    On stationary tests in the presence of structural breaks

    Econ. Lett.

    (1997)
  • C.-C. Lee et al.

    Stochastic convergence of per capita carbon dioxide emissions and multiple structural breaks in OECD countries

    Econ. Modell.

    (2009)
  • X. Li et al.

    Global convergence in per capita emissions

    Renew. Sustain. Energy Rev.

    (2013)
  • X.-L. Li et al.

    CO2 emissions converge in the 50 U.S. States-sequential panel selection method

    Econ. Modell.

    (2014)
  • S. Nazlioglu et al.

    A panel stationarity test with gradual structural shifts: Re-investigate the international commodity price shocks

    Econ. Modell.

    (2017)
  • U. Nyambuu et al.

    Climate change and the transition to a low carbon economy-carbon targets and the carbon budget

    Econ. Modell.

    (2020)
  • C. Ordas Criado et al.

    Convergence in per capita CO2 emissions: A robust distributional approach

    Resour. Energy Econ.

    (2011)
  • M.J. Presno et al.

    Stochastic convergence in per capita CO2 emissions: An approach from nonlinear stationarity analysis

    Energy Econ.

    (2018)
  • V. Rios et al.

    Convergence in CO2 emissions: A spatial economic analysis with cross-country interactions

    Energy Econ.

    (2018)
  • D. Romero-Avila

    Convergence in carbon dioxide emissions among industrialized countries revisited

    Energy Econ.

    (2008)
  • Y. Wang et al.

    Convergence behavior of carbon dioxide emissions in China

    Econ. Modell.

    (2014)
  • J. Wu et al.

    Convergence of carbon dioxide emissions in Chinese cities: A continuous dynamic distribution approach

    Energy Pol.

    (2016)
  • J. Yang et al.

    Carbon dioxide emissions and interregional economic convergence in China

    Econ. Modell.

    (2016)
  • S. Acar et al.

    Convergence of per capita carbon dioxide emissions: Implications and meta-analysis

    Clim. Pol.

    (2018)
  • J.E. Aldy

    Per capita carbon dioxide emissions: convergence or divergence?

    Environ. Resour. Econ.

    (2006)
  • J. Bai et al.

    Determining the number of factors in approximate factor models

    Econometrica

    (2002)
  • Cited by (0)

    Authors would like to thank the editor (S. Mallick), associate editor, and two anonymous referees for their helpful comments to improve the paper. All errors remain our own.

    1

    Saban Nazlioglu gratefully acknowledges that his work is a part of the project supported by The Scientific and Technological Research Council of Turkey (TUBITAK) under grant number 215K086.

    View full text