انتشار دی اکسید کربن و رشد اقتصادی در ایالات متحده

The objective of this paper is to analyze the relationship of the carbon Kuznets curve. We discuss two potential flaws in past carbon Kuznets curve studies: one, the potential misspecification of energy consumption as a control variable; and, two, the use of vector error correction models as an empirical specification. Given these potential flaws we estimate a dynamic ordinary least squares model of monthly carbon dioxide emissions, personal income, and energy production in the U.S. from 1981 to 2003. Our results suggest that economic growth drives emissions intensities, not absolute emissions as is often implied in past studies.

The objective of this paper is to integrate the approaches of the environmental Kuznets curve (EKC), decomposition analysis, and energy-growth based on three relatively independent streams of literature into one cohesive discussion on the nexus among energy, the environment, and the economy. Developing a discussion will aid in our understanding of energy resources as one of the major factors affecting both the economy and the environment. With current discussions surrounding global climate change, greenhouse gas emissions (GHGs), and economic development, the results within this study imply that economic development (including energy infrastructure development) can be beneficial to the environment (and arguably is necessary for environmental quality improvements). The environmental benefits of economic development as suggested by the EKC hypothesis are not to discredit those that advocate for sustainable growth, but rather to further the energy-pollution-income debate so that the public is aware of policies alternative to the limits-to-growth argument. More importantly, energy policy could play a significant role in stimulating economic development while at the same time ameliorating harmful pollution emissions. This study offers a unique contribution to the literature in four ways. The first contribution involves an empirical example in which we synthesize the three separate literatures outlined below. The second contribution is noting the potential flaw in many past studies that have empirically estimated the effects of energy use and economic growth on carbon dioxide emissions. Namely, carbon dioxide emissions in many past studies are estimated based upon energy consumption within a particular jurisdiction. Several studies then proceed to specify a covariate of energy consumption on the right-hand side (RHS) of the model, which may lead to spurious empirical results due to simultaneity bias. The third contribution is to point out the potential flaw in specifying a dynamic econometric model, such as a vector autoregressive (VAR) model or vector error correction model (VECM), which are popular in many recent studies. Subject to these potential flaws we estimate a dynamic ordinary least squares (DOLS) model among monthly time series of U.S. personal income, CO2 emissions (based upon U.S. monthly energy consumption), and energy production from 1981 to 2003. Our fourth contribution constitutes the use of high frequency, monthly data from 1981 to 2003. Many past studies use annual data that miss potential inter-annual variation in the time series variables. Further, the higher frequency data provide better asymptotic qualities for dynamic regression analysis. Using this uniquely high frequency of data we find a long-run relationship between the variables that suggests the inverted-U shaped relationship between CO2 emissions and personal income as espoused by the EKC. However, based upon our estimated data for CO2 emissions, our results may suggest that this relationship is instead between CO2 emissions intensities and income.

In this paper we discussed and then synthesized the three separate literatures of the EKC, decomposition methods, and energy-growth. Our synthesis involved an examination of the relationship among CO2 emissions, energy production, and national income in the U.S. over the period 1981–2003 using a dynamic ordinary least squares model. The empirical results suggest the existence of a long-run relationship between the variables that is consistent with the environmental Kuznets curve hypothesis – these results are similar to Ang (2007). To complement the findings we used the IPAT model to decompose CO2 emissions into the factors of population growth, economic growth, and technological growth. The IPAT analysis seems consistent with “weak” form of the EKC but seems to contradict the EKC hypothesis in the “strong” form as defined by Roca and Alcantara (2001). Due to how CO2 emissions are estimated, our regression results may capture the relationship between CO2 emission intensities and income (Itkonen, 2012), which is consistent with the IPAT decomposition. These findings may suggest to policy makers that economic growth is relieving some pressure on the environment in terms of carbon intensities in the U.S. but not absolute emissions. This is consistent with what we observe in Fig. 1 as we observe an initial increasing trend in CO2 emission intensities followed by a leveling (or probable decreasing) of intensities over this period. This would suggest that perhaps an economic stimulus policy may be an effective approach in ameliorating but perhaps not completely mitigating CO2 emissions. Therefore, our findings seem to suggest that conservation policies (e.g., energy conservation policies) may need to be coupled with economic stimulus policies to mitigate emissions as a whole. In other words, our results suggest that income in the form of GDP is not sufficient in and of itself to relieve the pressure of carbon dioxide emissions. We outlined several problems, within the carbon Kuznets curve literature, that have practical implications for climate change mitigation policy. First and foremost, the CKC is based upon a reduced-form specification so it is difficult to infer causality or make policy prescriptions concerning the relationship among economic growth, energy demand, and environmental degradation. Moreover, we outlined several problems within the literature that merit further attention. One of the principle problems is in regards to how carbon dioxide emissions are estimated – because these emissions are estimated from energy use one needs to be careful in the interpretation of econometric models which use energy consumption as an explanatory variable. The interpretation of coefficients which are consistent with the EKC hypothesis needs to interpreted with caution as to whether the results can be interpreted as an actual decrease in emissions or a decrease in intensity. Further, statistical evidence of a decrease in emissions may be sensitive to the type of times series model selected as we pointed out the potential problem with the stability conditions of such models which use a quadratic specification. In the future, a global reduction in CO2 emissions is going to be based upon multilateral agreements between nations. The connection between economic growth, energy, and emissions will have a significant impact on the political economy of negotiating such agreements. Therefore, as future studies seek to determine the connection between the economy, energy, and emissions it is important to be cognizant of the potential issues pointed out in this study. That is, future studies need to exercise caution in how carbon dioxide emissions are measured and how to interpret decreases in emissions. Any such findings of decreases are sensitive to the chosen explanatory variables and time series model selected.