John M. Mutua Thomas Sterner Martin Borjesson
NCDE 2009, Oscarsborg, Norway 18-19th June 2009
Introduction and Problem
Methodology
Empirical Results
Conclusions and Policy Recommendations…….
Energy is key to economic growth and quality of life
Energy consumed in transport, power generation, manufacturing sector, agriculture…..
Over 120,000 units of new vehicles added on the Kenyan roads in 2008 -Total vehicular population > 800,000.
Increased consumption of petroleum products (In Kenya: 3.7 Billion Litrers-Projected to be 13.9 Billion by 2030)
Rising emissions of carbon dioxide (CO2), nitrous oxides (NOx), dinitrogen oxide (N2O), sulphur dioxide (SO2), volatile organic compounds (VOCs), some lead and particulate matter (PM).
Changing fuel prices, new energy and environment policy
initiatives have heightened interests in the appropriate level of fuel taxation.
70.0 21.0
9.0
0 10 20 30 40 50 60 70 80
Woodfuel Petroleum
Electricity Share of National
Energy Consumption Matrix (%)
Share of National
Energy Consumption Matrix (%)
Cost Fractions of Petroleum Product Prices
(percentage), September 2005
54.4
0.3
33.9
2.9
8.5
59.7
0.4
29
3.5
7.3 Landed Cost (Ex. KPRL -
per litre)
KOSF
Total Taxes & Levies
Fees & Transport Charges
Margin
Premium Motor Spirit Diesel
This paper seeks to establish the distributional effects of transport fuel taxes in Kenya. Are they regressive or progressive?
Aims to provide policy recommendations on distributional aspects of transport fuel taxes
The methodology of this study is in three parts.
First, it computes budget shares of transport in household expenditures for each category of
population classified by household income deciles.
Secondly develops Lorenz curves/Gini coefficients
lastly, computes Suit Indices following Suits (1977)
The budget shares are calculated as follows:
Where;
E=expenditure share of each decile
TE=household transport expenditure for each decile
THE=total household expenditure for each decile
THE Eid TE
To measure the progressivity of a tax, a figure similar to Lorenzo curve, but one in which the accumulated percent of tax burden is plotted
vertically against the accumulated percent of income on the horizontal axis.
) /
( 1
/ )
( K L K L K
S
dy y
T
L
x 100 x( )
0
) /
(
1 L K
S
x
x0 10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 70 80 90 100
Accumulated Percent of Total Income
Accumulated Percent of Tax Burden
X Y
0 10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 70 80 90 100
Accumulated Percent of Households
Accumulated Percent of Total Income
it is evident that the burden is lowest for low income households.
X Y
0 10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 70 80 90 100
Accumulated Percent of Total Income
Accumulated Percent of Tax Burden
Tax Burden from Private Transport
Tax Burden from Private and Public Transport Y=X
Z’
Z
X Y
(a). Tax progressivity Private Transport Taxes
Type of Tax Suit Index(S)
Tax calculated based on transport expenditures
0.483 Tax calculated based on transport income 0.464
(b) Tax progressivity for combined Private and Public transport fuel Taxes
Tax calculated based on transport expenditures
0.225 Tax calculated based on transport income 0.171
The study indicates that a tax on or gasoline is progressive over the bottom half of the income distribution
This is because many of the lower income households do not own any vehicles and a price increase would make poorer households reduce their driving
distance more than wealthy households
The study states that greater price responsiveness among low-income
households enhances the degree of progressivity in the lower-income groups while mitigating the degree of regressivity in the upper-income groups.
However Suits (1977) has noted that NO generally accepted index on how regressive or progressive a particular tax is.
The gini coefficient varies from +1 at the extreme progressivity where the
entire tax burden is borne by members of the highest income bracket, through zero for a proportional tax, to -1 at the extreme of regressivity at which the entire tax burden is borne by members of lowest income brackets.
The Suit Indices for Kenya were
0.464 for private transport
0.171-for a combined private and public transport
Increasing vehicular population due to increased
preference for private transport as a result of improved incomes and resultant wealth effect among the middle and upper class strata of their population.
Estimation of distributional effects of fuel taxes by expenditure and income deciles
Presented a Lorenz curve of Kenya household income, calculated tax burdens and suit indices for both private transport fuel taxes
Combined distributional effects of transport fuel taxes for both private and public transport fuels.
Firstly, there is need to improve the public transport system and encourage mass transit so as to reduce private ownership of vehicles and gasoline consumption. This could be through improvement in the railway system and public bus/metro
system.
Need to revise taxes on high gasoline consumption vehicles which are not used for public transport. This will reduce per capita consumption of gasoline and hence achieve abatement.
Examine revenue potential from gasoline taxes and evaluate how these taxes can be used to compensate citizens from welfare losses by improving service delivery in roads, both railways and road transport and health sectors among others.
Thank you very much (tusen takk )
