7. Eliminate the variable with the lowest t-statistic, run the tests. If the current specification fails any of the tests, the last variable eliminated is replaced, and the current specification is re-estimated eliminating the variable with the next lowest insignificant t-statistic.
8. If the current specification passes all tests, return to 7.
9. The process of variable elimination ends when a current specification passes the tests and either has all variables significant or cannot eliminate any remaining insignificant variable without failing one of the tests. Record the resulting terminal specification.
10. If there is any combination of explanatory variables different from those considered, choose it and go to 2.
11. Once all search paths have ended in a terminal specification, the final specification is chosen through a sequence of encompassing tests. If more than one specification passes, the specification with the minimum Schwarz criterion is the final specification. If no model passes, reopen the search on the non-redundant joint model (including testing against the general specification) using only a single search path and take the resulting model as the final specification.
12. The final specification is the intersection of the two specifications from each sub-sample.
Caution should be taken in the part of possible multicollinearity. For example, various indexes, such as those related to economic freedom or reforms development, are built on the basis of macroeconomic indicators, so they should not appear in the same regression. The panel is likely to be unbalanced. If this is really so appropriate estimation procedure is applied [7, p. 159 – 164].
__________
References: 1. Levine, R. and D. Renelt. A Sensitivity Analysis of Cross-country Growth Regressions // American Economic Review 82. — 1992. — №4. — P. 942 – 963. 2. Sala-i-Martin, X. I Just Ran 2 Million Regressions // American Economic Review, 1997a. 3. Sala-i-Martin, X. I Just Ran Four Million Regressions // NBER Working Paper, 1997b. — №6252. 4. Doppelhofer, G., R. I. Miller, and X. Sala-i-Martin. Determinants of Long-Term Growth: A Bayesian Averaging of Classical Estimates (BACE) Approach // NBER Working Paper, 2000. — №7750. 5. Hoover, K. D. and S. J. Perez. Evaluating Specification Search in Cross-country Growth Regressions: Extreme Bounds vs. General-to-Specific Methodologies. Revised, 2001. 6. Hoover, K. D. and S. J. Perez. Truth and Robustness in Cross-country Growth Regressions // Journal of Economic Literature, 2000. 7. Baltagi, B. H. Econometric Analysis of Panel Data, 2001.
|
УДК 339.9 |
Piskunov K. |
Estimation of Costs of the Renewable Energy Sources Sector Development
under the EU Requirements
One of the strategic goals of Ukraine may be described as an entry into the European Union. Among the general criteria for eligibility is the question of Ukraine’s capacity to fulfill the obligation of membership, i. e. conformation to the legislation and the policies of the EC.
One of the most important requirements concerning the energy sector is the promotion of RES (renewable energy sources) of electricity production. In September, 2001, the EC Commission set the target of achieving by 2010 the 12% share of total consumption of electricity to be produced from RES. Hence, each member-country is expected to increase the production of electricity from RES in accordance with its national commitments under the Kyoto Protocol (table) [1].
RES, as stated in the Directive 2001/77/EC of the European Parliament and of the Council of 27 September, 2001 [1], include all the non-fossil energy sources, e. g. wind, solar, geothermal, wave, tidal ones, hydropower, biomass, landfill gas, sewage treatment plant gas and biogases.
Уважаемый посетитель!
Чтобы распечатать файл, скачайте его (в формате Word).
Ссылка на скачивание - внизу страницы.
