RESEARCH ARTICLE An investigation on the determinants of carbon emissions for OECD countries: empirical evidence from panel models robust to heterogeneity and cross-sectional dependence Eyup Dogan 1 & Fahri Seker 2 Received: 24 December 2015 /Accepted: 4 April 2016 # Springer-Verlag Berlin Heidelberg 2016 Abstract This empirical study analyzes the impacts of real income, energy consumption, financial development and trade openness on CO 2 emissions for the OECD countries in the Environmental Kuznets Curve (EKC) model by using panel econometric approaches that consider issues of hetero- geneity and cross-sectional dependence. Results from the Pesaran CD test, the Pesaran-Yamagata’ s homogeneity test, the CADF and the CIPS unit root tests, the LM bootstrap cointegration test, the DSUR estimator, and the Emirmahmutoglu-Kose Granger causality test indicate that (i) the panel time-series data are heterogeneous and cross- sectionally dependent; (ii) CO 2 emissions, real income, the quadratic income, energy consumption, financial develop- ment and openness are integrated of order one; (iii) the ana- lyzed data are cointegrated; (iv) the EKC hypothesis is vali- dated for the OECD countries; (v) increases in openness and financial development mitigate the level of emissions whereas energy consumption contributes to carbon emissions; (vi) a variety of Granger causal relationship is detected among the analyzed variables; and (vii) empirical results and policy rec- ommendations are accurate and efficient since panel econo- metric models used in this study account for heterogeneity and cross-sectional dependence in their estimation procedures. Keywords Carbon emissions . EKC model . Heterogeneity . Cross-sectional dependence . OECD Introduction The last annual conference of the United Nations Framework Convention on Climate Change took place in Paris in November–December 2015. 1 As a result of extraordinary in- creases in environmental pollution and global warming, 196 nations that participated at the Paris meeting have agreed to keep the global warming below 2 °C above pre-industrial level. The achievement of this target requires continuous ex- treme declines in the level of greenhouse gas emissions. Because some of the largest emitters of greenhouse gases— the USA, India, Russian Federation, Japan and Canada—have not signed the second amendment of the Kyoto protocol which aims to decline the levels of world emissions through bindings for parties to the protocol; 2 it is for now doubtful whether or not the aforementioned largest emitters are going to ratify the Paris agreement. Nevertheless, given the fact that environmental pollution is currently a real and urgent concern to handle, researchers and policy makers from many countries are willing to know possible determinants of greenhouse gas emissions. Studies including Soytas et al. (2007), Lean and Smyth (2010), Yavuz (2014), Shahbaz et al. (2014), Dogan et al. (2015), Kasman and Duman (2015), Seker et al. (2015), Al- Mulali et al. (2015), Tang and Tan (2015), Farhani and Ozturk (2015), Dogan and Turkekul (2016), Javid and Sharif (2016), 1 http://unfccc.int/resource/docs/2015/cop21/eng/l09r01.pdf (accessed 16 December 2015). 2 http://unfccc.int/kyoto_protocol/doha_amendment/items/7362.php (accessed 16 December 2015). Responsible editor: Philippe Garrigues * Eyup Dogan eyup.dogan@agu.edu.tr Fahri Seker fahri.fseker@gmail.com 1 Department of Economics, Abdullah Gul University, Sumer Campus, Office #B211e, Kayseri, Turkey 2 Department of Economics, Bozok University, Yozgat, Turkey Environ Sci Pollut Res DOI 10.1007/s11356-016-6632-2