Assessment of agricultural carbon emission performance and stochastic con-vergence in China using SBM-Undesirable model and panel unit root test

On overall scale and industrial perspective, researches on the evaluation of carbon emissions and the test of convergence in China have matured. However, the study has remained relatively weak in agricultural carbon emissions. To complement not only existing research but also for better understanding of the carbon emissions, performance and convergence in different regions, the paper used SBM-Undesirable model to assess the performance of agricultural carbon emissions based on the estimation of agricultural carbon emissions in 30 provinces in 2000-2014. Then three panel unit root tests were selected to determine stochastic convergence test for the investigated districts. The main conclusions were as follows:1) for the period 2000-2014, the overall trend in national agricultural carbon emissions increased, but the quantities of the emissions in different regions were significantly different. The mean agricultural carbon emission in the middle region was much larger than that in the whole nation, the eastern region and western region. The gaps in agricultural carbon emissions between the middle region and the whole nation, the eastern and western region were 3.357 4×10⁶ t, 3.965 0×10⁶ t and 5.904 7×10⁶ t respectively in 2000, whereas this gaps widened to 5.244 8×10⁶ t, 7.351 2×10⁶ t and 7.681 0×10⁶ t in 2014, corresponding respectively to growth rates of 56.2%, 85.4% and 30.0%. 2) The performance of agricultural carbon emissions in different regions turned out to differ apparently from distinct to district. A line graph of the average performance suggested that the performance was better for the eastern region, which was stable at 0.8 for 15 years. On the contrary, the average performance was relatively low for the west and middle regions, which was for most of the time within 0.3-0.5. The performance improved for the western region. However, the trend for the middle region was apparently the reverse. 3) In terms of convergence test of quantity, the examination of simulated convergence confirmed that stochastic convergence occurred only for the eastern region. There was no sign that stochastic convergence existed for the whole country, western region or even middle region. In the test of performance, there was no stochastic convergence for the whole country, while three regions exhibited relatively obviousness in the trend in club convergence. The results suggested that neither the quantity nor the performance of the whole country was automatically reducible to steady-state level. Thus it was necessary to make effective policy intervention to narrow the gap among the regions. Finally, this paper provided a further data-driven reference base for developing reasonable policies for the reductions of regional and national carbon emissions.