Skip to main content
Log in

A smooth coefficient model of carbon emissions

  • Published:
Empirical Economics Aims and scope Submit manuscript

Abstract

I use a semiparametric smooth coefficient model to estimate a generalization of the emissions convergence models derived from the green Solow model proposed by Brock and Taylor (J Econ Growth 15:127–153, 2010). Parametric estimates of simple homogeneous coefficient convergence models suggest that there may be heterogeneity in emissions convergence across different subsamples of observations. The semiparametric models confirm that there is heterogeneity across countries in coefficient estimates; however, such heterogeneity does not appear to be substantial enough to qualitatively influence the estimates derived from the parametric models. Hence, I find that (i) the green Solow model is a robust framework for analyzing carbon emissions convergence and (ii) carbon emissions are converging across a large sample of countries. My results suggest that international agreements that assign pollution rights based on population levels may be agreeable to many nations.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

Notes

  1. Aldy (2006), for example, notes the high correlation between carbon emissions and income.

  2. Perhaps a normative climate change perspective would be to encourage low emissions nations to maintain a low level of emissions, while high emissions nations reduce their emissions levels, i.e., convergence to a low level of emissions. In general, evidence of convergence does not necessarily imply that the steady state level of emissions to which the countries are converging is relatively low or high.

  3. There exist several other papers focusing on convergence in carbon emissions, albeit not in the absolute or conditional beta-convergence sense. Westerlund and Basher (2008) and Panopoulou and Pantelidis (2009), for example find evidence of per capita carbon emissions convergence using panel unit root tests and a nonlinear time-varying factor model, respectively. McKibbin and Stegman (2005), Stegman (2005), Aldy (2006), and Barassi et al. (2008), for example, do not find evidence of carbon convergence.

  4. The conditional variables in the conditional convergence equation of Strazicich and List (2003) include the natural logarithm of per capita GDP and its square, the average price of gasoline, the population density measured in 1978, and the average winter temperature.

  5. Durlauf et al. (2001) make a similar generalization of a standard Solow specification in the GDP growth context.

  6. It is important to recognize two limitations of the green Solow model. The first is that parameters in a reduced form specification such as the convergence equations discussed below are not policy invariant and should be interpreted with such an insight in mind (Lucas 1976). The second is the debate about whether the estimated convergence parameter in beta-convergence regressions can be informative about the dynamics of the distribution of emissions; see Friedman (1992), Quah (1993), and Bliss (1999) for important discussions. See, also, Van (2005) for an important empirical assessment of the dynamics of the world distribution of carbon emissions.

  7. It is important to note that Brock and Taylor (2010) argue that the \(\log [1-\theta _i]\) term in the conditional convergence equation is zero for per capita carbon emissions measured prior to 1998.

  8. In later sections, I define \(Z\) to be a matrix that includes both country and time indicators. In these models, I use the ordered discrete kernel function given by

    $$\begin{aligned} K(z,z_i) = \left\{ \begin{array}{ll} 1&\quad \text{ if}\quad z = z_i\\ h^{|z-z_i|}&\quad \text{ if}\quad z \ne z_i \end{array} \right. \end{aligned}$$
    (12)

    to smooth the time indicator, and note that \(K(z,z_i)\) becomes a product kernel function defined as the product of kernel weights for both the country and time indicators.

  9. The semiparametric smooth coefficient model has previously been applied by Durlauf et al. (2001), Mamuneas et al. (2006), and Hartarska et al. (2011) to name a few.

  10. In a cross-sectional regression, each country has only one observation, so a discussion of a bandwidth equal to zero implies regression using only one datapoint and hence does not make much sense. However, it is important to note that \(h \rightarrow 0\) as \(n \rightarrow \infty \) and \(nh \rightarrow \infty \) is a requirement for consistency in semiparametric and nonparametric kernel estimation.

  11. Note that the sample size is larger for the cross-sectional regressions in the first three columns of Tables 4 and 5 than for the primary cross-sectional sample. Regressions run using the original 158 countries yield qualitatively identical results to those in Tables 4 and 5.

  12. The bandwidths for the year indicator are 0.6328 and 0.7872 for the absolute and conditional convergence models, respectively.

References

  • Aldy JE (2006) Per capita carbon dioxide emissions? Convergence or divergence? Environ Res Econ 33:533–555

    Article  Google Scholar 

  • Barassi MR, Cole MA, Elliott RJR (2008) Stochastic divergence or convergence of per capita carbon dioxide emissions: re-examining the evidence. Environ Res Econ 40:121–137

    Article  Google Scholar 

  • Bliss C (1999) Galton’s fallacy and economic convergence. Oxf Econ Pap 51:4–14

    Article  Google Scholar 

  • Brock WA, Taylor MS (2010) The green Solow model. J Econ Growth 15:127–153

    Article  Google Scholar 

  • Durlauf SN, Kourtellos A, Minkin A (2001) The local Solow growth model. Eur Econ Rev 45:928–940

    Article  Google Scholar 

  • Friedman M (1992) Do old fallacies ever die? J Econ Lit 30:2129–2132

    Google Scholar 

  • Hartarska V, Parmeter CF, Nadolynak D (2011) Economies of scope of lending and mobilizing deposits in microfinance institutions: a semiparametric analysis. Am J Agric Econ 93:389–398

    Google Scholar 

  • Henderson DJ, Kumbhakar SC, Parmeter CF (2012) A simple method to visualize results in nonlinear regression models. Econ Lett 117:578–581

    Google Scholar 

  • Heston A, Summers R, Aten B (2009) Penn World Table Version 6.3, Center for International Comparisions of Production, Income and Prices at the University of Pennsylvania

  • Hurvich CM, Simonoff JS, Tsai CL (1998) Smoothing parameter selection in nonparametric regression using an improved Akaike information criterion. J R Stat Soc Ser B 60:271–293

    Article  Google Scholar 

  • Li Q, Racine JS (2004) Cross-validated local linear nonparametric regression. Stat Sin 14:485–512

    Google Scholar 

  • Li Q, Racine JS (2010) Smooth varying coefficient estimation and inference for qualitative and quantitative data. Econom Theory 26:1607–1637

    Article  Google Scholar 

  • Li Q, Huang CJ, Li D, Fu T (2002) Semiparametric smooth coefficient models. J Bus Econ Stat 20:412–422

    Article  Google Scholar 

  • Li Q, Ouyang D, Racine JS (2011) Categorical semiparametric varying coefficient models. J Appl Econom. doi:10.1002/Jae.1261

  • Lucas RE (1976) Econometric policy evaluation: a critique. In: Brunner K, Meltzer AH (eds) The Phillips curve and labor markets. Carnegie-Rochester conference on public policy. North Holland, Amsterdam

  • Mamuneas TP, Savvides A, Stengos T (2006) Economic development and the return to human capital: a smooth coefficient semiparametric approach. J Appl Econom 21:111–132

    Article  Google Scholar 

  • Mankiw NG, Romer D, Weil DN (1992) A contribution to the empirics of economic growth. Q J Econ 107:407–437

    Article  Google Scholar 

  • McKibbin W, Stegman A (2005) Convergence and per capita carbon emissions. Working paper

  • Meyer A (2006) The fair choice for climate change. BBC News. http://news.bbc.co.uk/2/hi/science/nature/4994296.stm. Accessed 11 June 2012

  • Panopoulou E, Pantelidis T (2009) Club convergence in carbon dioxide emissions. Environ Res Econ 44:47–70

    Article  Google Scholar 

  • Quah D (1993) Galton’s fallacy and tests of the convergence hypothesis. Scand J Econ 95:427–443

    Article  Google Scholar 

  • Racine J, Li Q (2004) Nonparametric estimation of regression functions with both categorical and continuous data. J Econom 119:99–130

    Article  Google Scholar 

  • Simonoff JS, Tsai CL (1999) Semiparametric and additive model selection using an improved Akaike information criterion. J Comput Graph Stat 8:22–40

    Google Scholar 

  • Strazicich MC, List JA (2003) Are \(\text{ CO}_2\) emissions levels converging among industrial countries? Environ Res Econ 24:263–271

    Article  Google Scholar 

  • Stegman A (2005) Convergence in carbon emissions per capita. Working paper

  • Van PN (2005) Distribution dynamics of \(\text{ CO}_2\) emissions. Environ Res Econ 32:495–508

    Article  Google Scholar 

  • Westerlund J, Basher SA (2008) Testing for convergence in carbon dioxide emissions using a century of panel data. Environ Res Econ 40:109–120

    Article  Google Scholar 

  • Zhang R, Sun K, Delgado MS, Kumbhakar SC (2012) Productivity in China’s high technology industry: regional heterogeneity and R&D. Technol Forecast Soc Change 79:127–141

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Michael S. Delgado.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Delgado, M.S. A smooth coefficient model of carbon emissions. Empir Econ 45, 1049–1071 (2013). https://doi.org/10.1007/s00181-012-0658-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00181-012-0658-1

Keywords

Navigation