9

It is well known that during the past two decades the developed countries have madeimportant efforts in reducing emissions of pollutants. However, as the results in Table Isuggest, these have not been equally important for all OECD countries. In this regard, theclassification of countries into those that performed better than average and those belowthe average with respect to the overall environmental performance indicator sheds lightson the relative importance attached to environmental concerns while pursuing growthobjectives. A comparison of the average indicators reveals that, relatively low income countries (asmeasured by average per-capita GDP between 1971-1990 expressed in internationalprices (1985 U.S. dollars)) while trying to catch-up with relatively high income countriesmay have ignored environmental concerns. Note that while these countries achievedhigher growth rates for the good output than high income countries, their emission ofpollutants have increased at an even faster rate, lowering their environmentalperformance below 1971 levels. The relatively high income countries are the ones whichachieved lower than average growth in good output with rather low (and in 5 cases withnegative) growth rates in bad output (again evaluated with respect to the base year 1971).This is an indication that environmental concerns are becoming a binding constraint onlyafter a certain level of income is reached and this can be best addressed in anEnvironmental Kuznets Curve context using panel data, to which we turn next.Please insert Table I hereLetting itEPrepresent the environmental performance of country iin year t, the equationbelow specifies a possible relation between environmental performance and per-capitaGDP (GDPPC).ititititiitGDPPCGDPPCGDPPCEP342321)()(where: i: country index; t: time index; : disturbance term with mean zero and finitevariance.10

The shape of the polynomial will reveal the relationship between environmentalefficiency and GDP per capita. A negative sign for GDPPC coupled with a positive signfor its quadratic and a negative sign for its cubic terms will imply deterioratingenvironmental performance at the initial phases of growth which is followed by a phaseof improvement and then a further deterioration once a critical level of per capita GDP isreached. Models that combine cross-section and time-series data rely on the premise thatdifferences across units can be captured in differences in the intercept term. Howeverestimation techniques differ with respect to the nature of assumptions made on theintercept of the equation. If the 1iare assumed to be fixed parameters, then the modelis known as a fixed effects model. If on the other hand the 1iare assumed to be randomvariables that are expressed as 11ii, where 1is an unknown parameter and iare independent and identically distributed random variables with mean zero and constantvariance, then the model is called a random effects model. The disadvantage of the fixed

#### You've reached the end of your free preview.

Want to read all 27 pages?

- Spring '14
- DanielKevles
- Environmental Economics, Per capita income, Random effects model, environmental performance, Environmental Performance Index