You are on page 1of 10

ENVIRONMENTAL ECONOMICS

& POLICY
EKC
EKC
e = emissions
per capita
e e
Stage Stage Stage
e  y 1 2 3

e   0 y  1 y 2

y=GDP per capita    0  1 y y y

Emissions strictly EKC hypothesis Panayotou Relation


increasing
EKC
 Empirical Evidence:

 Original study by Grossman + Krueger (1991) on NAFTA found EKC for SO2,
Suspended Particulate Matter and Fine Smoke.

 Shafik + Bandopadhyay (1992) use World bank data and find evidence of EKC for
SO2 and Suspended Particulate Matter, but find none for CO2 or deforestation.

 Hettige et al. (1992) develop a toxic intensity index for manufacturing sectors in
developed + developing nations. They find NO EKC for toxic intensity per unit
manufacturing output, but toxic intensity falls with GDP  manufacturing does
not become cleaner, but the latter effect is driven by a change in tastes.

 Selden + Song (1994) find EKC for SO2, Suspended Particulates, Nitrogen Oxide
and Carbon Monoxide, although higher turning points for emissions than ambient
concentrations.

 Cole et al. (1997) using data for OECD countries found EKC for CO2, CO, Nitrates,
NO, SO2, Suspended Particulates.

 Cole (2003) conducted a robust study, using different functional forms, different
country samples and variables affecting income distribution. STILL FINDS EKC for
SO2, NO and even CO2.

 Panayotou (1995) finds an EKC for deforestation.


EKC
 The evidence appears to support a notion that an EKC – style relationship holds for pollutants with a
localised impact and where there are generalised costs and high potential private and social benefits.

 For pollutants with inter-country or inter-regional impacts (such as CO2), the relationship does not
appear to be present.

 Stern (2004) further suggests that the evidence indicates an EKC hypothesis exists for
CONCENTRATIONS OF POLLUTANTS, whilst emissions are still monotonically increasing in income.

 Part of this could be explained by the dispersion of industry as income increases:

 During early industrialisation, industrialisation tends to be highly clustered, and it decentralises with
time. Therefore, although total emissions may still increase, or stay level, but will become less
concentrated.

 Yandle et al. (2004) also note that there is nothing automatic about the EKC relationship observed in
OECD countries for select pollutants. Higher GDP growth needs to be supported by policies and
institutions, such as laws, regulation, environmental culture + norms and political pressure.

 Most importantly they argue that moving across the EKC is akin to moving through the evolution of
PROPERTY RIGHTS:

 Start with open access/commons, move through to public/communal property rights and finally to
private property rights, where the latter has the greatest incentive to prevent externalities due to a ‘not-
in-my-backyard’ effect.

 In net, the empirical results on the EKC do not suggest that purely growth-enhancing policies are a
suitable substitute for environmental regulation, or a laissez-faire attitude towards environmental
protection.
EKC
EKC

 Extensions to the original model:

 Stern (2004) notes three alternative variations on the traditional EKC curve (see
next slide for diagrams):

 The ‘New Toxics’ approach suggest that, while emissions of old pollutants fall
over time, new substitutes take their place; overall emissions therefore continue
to rise.

 The ‘revised EKC’ approach suggests the same relationship as the traditional EKC,
however, the curve shifts down and to the left over time due to technological
progress.

 The ‘Race to the Bottom’ approach suggests that emissions in developed


countries are reduced by outsourcing to developing nations, who will find it hard
to reduce emissions. This may prevent emissions from declining, although they
may flat line.

 Stern (2004) concludes that the ‘true’ EKC shape is that of a combination between
‘New Toxics’ and ‘Revised EKC’  the basic shape of the new toxics curve that
shifts down over time due to innovations.

 However, they also note that the relationship is likely to be different for different
groups of pollutants.
EKC
Variations on the EKC

Pollution

New Toxics

Race to the
Bottom

Trad. EKC
Revised EKC

Income per capita


EKC
 The long-run effect:

 One distinguish between two scenarios:

 A: Environmental impact per unit income tends to 0 as per capita income rises (OPTIMISTIC
 laws of thermodynamics suggest that waste will be created in production of goods).

 B: Environmental impact per unit income falls to a strictly positive level ‘k’ (MORE
REALISTIC).

 If we consider a globally binding EKC curve (which is unrealistic, due to the locality of EKC-
applicable pollutants), then a simple analysis can be made using two countries:

 Developed nation – initially has a greater income, grows at rate ‘g’.

 Developing nation – lags behind the developed nation, also grows at rate ‘g’.

 See next slide for diagrams; In scenario A, there is a ‘double hump’ due to the developing
nation hitting the turning point of income later than the developed nation. However,
emissions  0 as time  infinity.

 In scenario B, environmental impact per unit income  k as time  infinity. However, this
also means that TOTAL IMPACT is infinitely increasing.

 Therefore, the EKC may be true for certain pollutants in individual countries, but in the very
long run on a global scale it is unlikely to lead to the results postulated in the original model,
ceteris paribus.
EKC
e = emissions
per unit income
e e

y t t

Scenarios A and B Scenario A (Global) Scenario B (Global)

You might also like