Transcript GTAP Resource 4386

TOWARDS FINDING THE TAX
INCIDENCE OF CARBON TAXES IN
SOUTH AFRICA
Jan H van Heerden
Heinrich Bohlmann
OUTLINE OF THE PAPER
•
•
•
•
•
The Problem
Possible Solutions
Previous Study
The Data
Adjusting the Model
•
•
•
•
Policy Simulations
Results
Conclusion
Further work
THE PROBLEM
• South Africa ranks amongst the first world countries in
the world in CO2 pollution, and its footprint looks bad
CO2 per capita: 1999
25
20
15
10
5
Ki
ng
Cz
do
ec
m
h
Re
pu
bl
ic
Au
st
ra
Un
lia
it e
d
St
at
es
Ko
re
a
Un
it e
d
Po
la
nd
So
ut
h
Af
ric
a
Ita
ly
Po
rtu
ov
ga
ak
l
Re
pu
bl
ic
Sl
co
ex
i
M
In
di
a
0
Ton CO2 per capita
CO2/95 pppUS$ GDP: 1999
1,2
1
Emissions intensity
0,8
0,6
0,4
0,2
Ko
Un
re
a
it e
d
St
at
es
Au
st
Sl
ra
ov
lia
ak
Re
pu
Cz
bl
ec
ic
h
Re
pu
bl
ic
Po
la
nd
So
ut
h
Af
ric
a
co
ex
i
M
ng
do
m
Ki
In
di
a
Un
it e
d
Po
rtu
ga
l
Ita
ly
0
kg CO2/95 pppUS$ GDP
Source: International Environmental Agency (IEA). 2001. Key world energy
statistics. Paris: IEA. (www.iea.org/statist/key2001/keyworld-2001.pdf)
GHG Emissions M-Tons - 2009
Rank
Country
M - ton
%
1
China
7 711
25.40%
2
United States
5 425
17.80%
3
India
1 602
5.30%
4
Russia
1 572
5.20%
5
Japan
1 098
3.60%
6
Germany
766
2.50%
7
Canada
541
1.80%
8
Korea, South
528
1.70%
9
Iran
527
1.70%
10
United Kingdom
520
1.70%
11
Saudi Arabia
470
1.50%
12
South Africa
450
1.50%
13
Mexico
444
1.50%
14
Brazil
420
1.40%
15
Australia
418
1.40%
16
Indonesia
413
1.40%
17
Italy
408
1.30%
18
France
397
1.30%
19
Spain
330
1.10%
20
Taiwan
291
1.00%
21
Poland
286
0.90%
POSSIBLE SOLUTIONS
1. Carbon Emissions Tax
Actual measured emissions; or
2. Proxy tax bases:
A.
Fossil Fuel Input (Upstream):
where fuels enter the economy based on the carbon content of the
fuel.
B.
Output Tax (Downstream):
(i) At point where fuel is combusted.
(ii) May be based on average emissions of production
processes.
Previously
• In 2004/5 the Dutch government funded a project
(PREM) to search for double dividends in the
environment and economy of South Africa.
• We used a static CGE model to simulate the effects of
carbon, fuel and energy taxes in the country.
• We found triple dividends with some tax and recycling
combinations (environment, economy and poverty)
•
Van Heerden, et al., Searching for Triple Dividends in South Africa: Fighting
CO2 pollution and poverty while promoting growth, The Energy Journal,
2006
This paper
• Gives preliminary results of a World Bank project to
search for double dividends in the environment and
economy of South Africa.
• We use a dynamic CGE model to
• expand the electricity industry from being a single
producer and distributor of electricity to a few
generators and one distributor, and
• simulate the effects of a fuel input tax in the country.
THE DATA (1)
• Updated 2011 database of South Africa
• Core data taken from the 2011 SU tables (StatsSA)
• Database aggregated to 45 sectors, with the electricity
sector
then
split
between
8
generators
transmitter/distributor based on available data.
and
1
THE DATA (2)
Electricity
Supply, R
Leontief
Good 1
up to
(not electricity)
Good N
(not electricity)
Primary
factors
CES
CES
CES
Imported
Good 1
Domestic
Good 1
Imported
Good N
Electricity
Domestic
Good N
Land
CES
Good 1 from
region 1
Good 1 from
region 2
Other costs
Labour
Capital
CES
up to
Good 1 from
region R
Labour
type 1
Labour
type 2
NEM
CES
Generation 1,
NEM region 1
Generation M,
NEM region 1
Source: MMRF document from http://www.copsmodels.com/archivep.htm#tppa0080
up to
Generation M,
NEM region N
up to
Labour
type O
Database split of the electricity sector
• We used the procedure followed by the MMRF model of
CoPS:
• Database split.docx
THE MODEL (1)
• Change in revenue dR= T.dX + X.dT
• T is rate and X is base
• But % change in X is x = 100*dX/X
• Therefore dR = TxX/100 + X.dT
•
= Rx/100 + X.dT
• dR affects government revenue and dT all prices
THE MODEL (2)
•
! Leontief demand for inputs !
•
Equation E_x1_sa # Demands for commodity composites #
(all,c,COM)(all,i,IND52) x1_s(c,i) - [a1_s(c,i) + a1tot(i)] = z(i);
•
Equation E_x1_sb # Demands for commodity composites #
(all,c,COM45) x1_s(c,"ElecSup") - [a1_s(c,"ElecSup") + a1tot("ElecSup")]
= z("ElecSup");
! CES demand for inputs !
•
Equation E_x1_sc
(all,c,GEN) x1_s(c,"ElecSup") - a1_s(c,"ElecSup")
•
= z("ElecSup") - SIGMAGEN(c)*[p1_s(c,"ElecSup") + a1_s(c,"ElecSup") - p1_gen];
POLICY SIMULATIONS
• The modelling exercises focus on two pieces of
government policy in South Africa
• Integrated Resource Plan (IRP) for Electricity (2010-2030)
•
http://www.doe-irp.co.za/content/IRP2010_updatea.pdf
• Carbon tax of R120/ton CO2e from 2016
•
http://www.thedti.gov.za/parliament/Reducing_greenhouse_gas.pdf
Baseline forecast (1)
120
% change in selected macro-economic
variables (cumulative)
100
80
60
40
20
0
2012
2017
Real GDP
2022
2027
2032
CO2: CoalGen constrained
Baseline forecast
% change in output growth for different power
generation sources (cumulative)
140
120
100
80
60
40
20
0
-20
2012
CoalGen
2017
WindGen
2022
SolarPVGen
2027
SolarCSPGen
2032
GasGen
OtherGen
RESULTS:
Carbon tax/no recycling
1,400
Million ton CO2-equiv
1,200
1,000
800
600
400
200
0
2015
2020
Growth without constraints (range)
Required by science (range)
Baseline
CO2: No CoalGen Constraints
2025
2030
Baseline & Tax
2035
CONCLUSIONS
• Implementing a CES demand function for generated electricity by
the supplying industry causes a switch to green electricity but not
nearly enough. Currently the supplier merely uses coal generated
power much more efficiently and not enough substitution takes
place.
• The carbon tax by itself – especially with all the exemptions for the
first five years – is not enough. Regulation of coal generated power,
as well as pro-active stimulation of green generation together with
the tax will be necessary to reach the targets.