Economic growth and CO2 emissions: an investigation with smooth transition autoregressive distributed lag models for the 1800-2014 period in the USA

Küçük Resim Yok



Dergi Başlığı

Dergi ISSN

Cilt Başlığı


Springer Heidelberg

Erişim Hakkı



The study aims to combine the autoregressive distributed lag (ARDL) cointegration framework with smooth transition autoregressive (STAR)-type nonlinear econometric models for causal inference. Further, the proposed STAR distributed lag (STARDL) models offer new insights in terms of modeling nonlinearity in the long- and short-run relations between analyzed variables. The STARDL method allows modeling and testing nonlinearity in the short-run and long-run parameters or both in the short- and long-run relations. To this aim, the relation between CO2 emissions and economic growth rates in the USA is investigated for the 1800-2014 period, which is one of the largest data sets available. The proposed hybrid models are the logistic, exponential, and second-order logistic smooth transition autoregressive distributed lag (LSTARDL, ESTARDL, and LSTAR2DL) models combine the STAR framework with nonlinear ARDL-type cointegration to augment the linear ARDL approach with smooth transitional nonlinearity. The proposed models provide a new approach to the relevant econometrics and environmental economics literature. Our results indicated the presence of asymmetric long-run and short-run relations between the analyzed variables that are from the GDP towards CO2 emissions. By the use of newly proposed STARDL models, the results are in favor of important differences in terms of the response of CO2 emissions in regimes 1 and 2 for the estimated LSTAR2DL and LSTARDL models.


Anahtar Kelimeler

Nonlinear time series, Cointegration, LSTAR-ARDL, ESTAR-ARDL, Environmental economics, Environmental Kuznets curve


Environmental Science And Pollution Research

WoS Q Değeri


Scopus Q Değeri