| Literature DB >> 31422533 |
Fakhri J Hasanov1,2,3, Jeyhun I Mikayilov1,4,5, Shahriyar Mukhtarov6,7, Elchin Suleymanov8,9.
Abstract
This paper investigates the CO2 emissions-economic growth relationship in Kazakhstan for the period 1992-2013. Johansen, ARDLBT, DOLS, FMOLS, and CCR cointegration methods are used for robustness purpose. We start with the cubic functional form to rule out any misleading results that can be caused by misspecification. Although the estimation results suggest "U"-shaped relationship, the turning point of income is out of the period. It means that the impact of economic growth on CO2 is monotonically increasing in the long run indicating the Environmental Kuznets Curve (EKC) hypothesis does not hold for Kazakhstan. Moreover, we calculate that the income elasticity of CO2 is about unity. The paper concludes that the Kazakhstani policymakers should focus on less energy-intensive sectors as well as using more renewable energy in order to avoid higher pollution effects of economic growth. They may also set new policy regulations for CO2 reduction.Entities:
Keywords: CO2 emissions; Cointegration; EKC hypothesis; Economic growth; Kazakhstan
Mesh:
Substances:
Year: 2019 PMID: 31422533 PMCID: PMC6823315 DOI: 10.1007/s11356-019-06166-y
Source DB: PubMed Journal: Environ Sci Pollut Res Int ISSN: 0944-1344 Impact factor: 4.223
Review of CO2 studies for Kazakhstan
| Study | Sample | Country or region | Functional form | Econometric methodology | Data | Income elasticity | Shape of EIR |
|---|---|---|---|---|---|---|---|
| Tamazian and Rao ( | 1993–2004 | 24 transition economies including Kazakhstan | QLF | GMM | PD | 0.04–1.22 ln GDPa | IUS |
| Apergis and Payne ( | 1992–2004 | 11 CIS countries including Kazakhstan | QLF | FMOLS | PD | For panel with Russia: 1.55–2.96 ln GDP; For panel without Russia: 1.37–2.54 ln GDPb | IUS |
| Stolyarova ( | 1960–2008 | 93 countries including Kazakhstan | LLF | GMM | PD | Short-run elasticity: 0.3–0.79c | Not reported |
| Brizga et al. ( | 1990–2010 | 15 former Soviet countries Kazakhstan | LLF | Index decomposition analysis and OLS | TD | 0.86 | Kazakhstan IUS |
| Perez-Suarez and Lopez-Menendez ( | 1860–2012 | 175 countries including Kazakhstan | CLF | NLS | TD | Not reported | No specific pattern |
| Narayan et al. ( | 1960–2008 | 181 countries including Kazakhstan | Correlation coefficients are used | Cross-correlation estimate | TD | Not reported | Kazakhstan US |
| Al-Mulali et al. ( | 1980(1990)-2010 | 107 countries including Kazakhstan | QLF | DOLS | PD | 4.75–0.18 ln GDP for the group with Kazakhstand,e,f | IUS for the group with Kazakhstan |
| Erdoğan and Ganiyev ( | 1992–2013 | Central Asia including Kazakhstan | QFF | Fixed effect and random effect | PD | 0.13 | IUS |
| Ito ( | 2002–2011 | 42 developing countries including Kazakhstan | LLF | GMM and PMG | PD | GMM: 0.13 PMG: 0.34 | MI |
| Mitic et al. ( | 1997–2014 | 17 transitional economies including Kazakhstan | LLF | DOLS and FMOLS | PD | 0.35 | MI |
| Shuai et al. ( | 1960–2011 | 164 countries including Kazakhstan | QFF | OLS | TD, PD | 0.81–0.08 ln GDP | IUS |
| Akbota and Baek ( | 1991–2014 | Kazakhstan | QFF | ARDL | TD | IUS |
aFor different specifications, the coefficients are slightly different; hence, we took the average of obtained coefficients (results from Table 4 were used) and calculated the elasticity.
bThe mean of ln GDP is not provided; hence, only elasticity formula is calculated
cQLF is used, but, then, the squared term is excluded due to multicollinearity
dOnly the model with growth rates is used
eAuthors’ calculation is based on the results of that study
fStarted with CLF and concluded with LLF
EIR, emission–income relationship; ARDLBT, autoregressive distributed lag bound testing; FMOLS, fully modified ordinary least squares; DOLS, dynamic ordinary least squares; GMM, generalized method of moments; OLS, ordinary least squares; PMG, pooled mean group estimator; NLS, non-linear least square method; PD and TD, panel data and time series data; CIS, commonwealth of independent states; MI, monotonically increasing; US, U-shaped; IUS, inverted U-shaped; LLF, log-linear function; QLF, quadratic functional form in logarithm; CLF, cubic functional form in logarithm
Modified from Mikayilov et al. (2018)
The VAR residual diagnostics and cointegration test results
| Panel A: serial correlation LM testa | |||||
| Lags | LM statistic | ||||
| 1 | 17.63 | 0.35 | |||
| 2 | 11.22 | 0.80 | |||
| 3 | 23.37 | 0.10 | |||
| Panel B: normality testb | |||||
| Statistic | χ2 | d.f. | |||
| Skewness | 5.48 | 4 | 0.24 | ||
| Kurtosis | 6.07 | 4 | 0.19 | ||
| Jarque–Bera | 11.55 | 8 | 0.17 | ||
| Panel C: stability testc | |||||
| Modulus | Root | ||||
| 0.96 | 0.91 − 0.29i | ||||
| 0.96 | 0.91 + 0.29i | ||||
| 0.67 | 0.18 − 0.64i | ||||
| 0.67 | 0.18 + 0.64i | ||||
| Panel D: Johansen cointegration test summary | |||||
| Data trend | None | None | Linear | Linear | Quadratic |
| Test type | (a) No | (b) Only | (c) Only | (d) | (e) |
| Trace | 4 | 3 | 1 | 2 | 1 |
| Max-Eig | 2 | 2 | 1 | 1 | 1 |
| Panel E: Johansen cointegration test results for type ( | |||||
| Null hypothesis | |||||
| 59.85*** | 29.26* | ||||
| 30.59** | 16.91 | ||||
aThe null hypothesis in the serial correlation LM test is that there is no serial correlation at lag order h of the residuals
bSystem normality test with the null hypothesis of the residuals is multivariate normal
cVAR stability test results show that no roots of characteristic polynomial are outside the unit circle
*Rejection of null hypothesis at 10% significance level
**Rejection of null hypothesis at 5% significance level
***Rejection of null hypothesis at 1% significance level
χ2, Chi-squared; d.f., degree of freedom; C and t, intercept and trend; r, rank of Π matrix, i.e., number of cointegrated equations; λtrace and λmax, trace and max-eigenvalue statistics
Critical values for the cointegration test are taken from MacKinnon et al. (1999); estimation period 1994–2013
Fig. 1Time profile of the variables (in logarithmic form)
Descriptive statistics of the variables
| Variable | Minimum | Mean | Maximum | Standard deviation | Coefficient of variation (%) |
|---|---|---|---|---|---|
| − 4.85 | − 4.48 | − 4.11 | 0.24 | 5.36 | |
| 8.23 | 8.70 | 9.24 | 0.36 | 4.14 |
The UR test results
| Variable | The ADF test | The KPSS test | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Test value | Test value | |||||||||
| − 2.71 | x | x | 0 | 0.18** | x | x | ||||
| − 2.53 | x | x | 1 | 0.13* | x | x | ||||
| − 2.52 | x | x | 1 | 0.14* | x | x | ||||
| − 2.50 | x | x | 1 | 0.14* | x | x | ||||
| Δ | − 1.97** | x | 2 | 0.44* | x | |||||
| Δ | − 2.67* | x | 2 | 0.37* | x | |||||
| Δ | − 2.65* | x | 2 | 0.39* | x | |||||
| Δ | − 2.63* | x | 2 | 0.40* | x | |||||
Maximum lag order is set to two and optimal lag order (k) is selected based on the Schwarz criterion in the tests. The critical values for the ADF and KPSS tests are taken from MacKinnon (1996) and Kwiatkowski et al. (1992), respectively. Estimation period 1992–2013. None means that neither intercept nor trend is included in the test equation. Note that unit root test equation can include one of the following: intercept (C), intercept and trend (t), and none of them (None). x indicates that the corresponding option is selected in the equation
**Indicates rejection of the null hypotheses of unit root in the ADF and stationarity or trend stationarity in the KPSS at 5% significance level
*Indicates rejection of the null hypotheses of unit root in the ADF and stationarity or trend stationarity in the KPSS at 10% significance level
The long-run equations
| Panel A: Cubic functional form: | ||||
| 704.57 | 2.83 | |||
| (453.63) | (52.30) | (2.01) | ||
| Panel B: Quadratic functional form, when | ||||
| χ2 (1) = 1.71 [0.19] | 36.13 | 2.09 | ||
| (7.96) | (0.46) | |||
Values in bracket and parentheses are probability and standard errors, respectively. Estimation period 1994–2013
Long-run coefficients from the methods
| Methods | ||||
|---|---|---|---|---|
| Coef. (std. er.) | Coef. (std. er.) | Coef. (std. er.) | Coef. (std. er.) | |
| VEC | − 12.05** (4.13) | 0.71** (0.24) | 46.82** (17.66) | |
| ARDLBT | − 15.25*** (4.80) | 0.93*** (0.29) | 58.23** (27.24) | − 0.04* (0.02) |
| DOLS | − 15.96*** (4.95) | 0.98*** (0.30) | 60.40** (20.75) | − 0.05** (0.02) |
| CCR | − 19.21*** (3.68) | 1.20*** (0.21) | 72.39*** (15.83) | − 0.08*** (0.01) |
| FMOLS | − 19.40*** (3.33) | 1.22*** (0.19) | 73.15*** (14.31) | − 0.08*** (0.01) |
The dependent variable is co2. The pulse dummy variable of DP2009 is significant in the FMOLS and CCR while insignificant in the DOLS as the first two have static structure and the last one has a dynamic structure. In DOLS, we set the maximum lag and lead being one, which is consistent with what we selected in the VAR/VEC and ARDL estimations and prefer Schwarz criterion to select an optimal ones. Estimation period covers 1994–2013
Coef., coefficient; Std. Er., standard error
*Indicates significance level at 10%
**Indicates significance level at 5%
***Indicates significance level at 1%
The ARDLBT estimation and test results
| Panel A: Selected ARDL specification | |||||
| Panel B: Residual diagnostics and misspecification test results for | |||||
| Panel C: The cointegration test results for | |||||
| The sample | Significance level (%) | Pesaran et al. ( | Narayan ( | ||
| Low bound | Upper bound | Low bound | Upper bound | ||
| 1 | 4.99 | 5.85 | 6.43 | 7.51 | |
| 5 | 3.88 | 4.61 | 4.54 | 5.42 | |
| 10 | 3.38 | 4.02 | 3.77 | 4.54 | |
| Panel D: Long-run relation derived from | |||||
| 15.25 | 0.93 | 0.04 | |||
| (4.80) | (0.07) | (0.02) | |||
, denote Chi-squared statistics to test the null hypotheses of no serial correlation, no autoregressive conditioned heteroscedasticity, and no heteroscedasticity in the residuals; JBand F indicate Jarque–Bera and F statistics to test the null hypotheses of normal distribution and no functional misspecification respectively; FW is the F-value of testing the null hypothesis of in the Wald test. Critical values are taken from the case of unrestricted intercept and restricted trend, 2 regressors and 30 observations (see Pesaran et al. 2001 and Narayan 2005). Probabilities are in brackets and standard errors are in parentheses. The pulse dummy variable of DP09Q1 is included in the estimation. Estimation period 1994–2013