Issues in Domestic Petroleum Pricing in Oil

Download Report

Transcript Issues in Domestic Petroleum Pricing in Oil

Issues in Domestic
Petroleum Pricing in OilProducing Countries
Sanjeev Gupta
IMF
January 27, 2004
Introduction: Why study
domestic petroleum prices?

Many oil-producing countries keep domestic prices
below free market levels

Net oil importers in some cases do not fully reflect
international prices.

The resulting subsidies can be large and are
nontransparent

Few studies have quantified these subsidies for a wide
range of countries and examined their economic effects
2
Subsidy Estimates: Methodology

Subsidies can be broadly defined as the difference
between the reduced price of a good with government
support and the price of the good in the absence of such
support.

Subsidy = (M-P)C

M = “free market price” = W+D+T



W = world wholesale spot price
D = marketing + distribution + transport costs (importers
only)
T = general consumption taxes (VAT, etc.)

P = after-tax retail price

C = petroleum product consumption
3
Subsidy Estimates: Caveats

Marketing, distribution, and transport costs assumed
to be the same across countries and time periods;
introduces some error, esp. if pipeline constraints
create large marginal transport costs

However, transport costs are typically only about 2
percent of after-tax retail price

Excludes implicit subsidies due to tolerance of
nonpayments

Some retail price estimates are an average for only a
month or quarter (“snapshot”)
4
Subsidy Estimates: Sample

Subsidies calculated for 86 countries between
1996-2000

1999 is the latest year available for most
countries. In this year, 62 countries in sample:



15 Major Oil Exporters (of which, 13 subsidize)
6 Other Net Oil Exporters
41 Net Oil Importers
5
Table 1. Domestic Petroleum Price Subsidies, 1999
(Median values in parentheses)
Average Oil Subsidies
(In percent of
government
expenditure)
(In percent of GDP)
Major Oil Exporters
Of which:
Subsidizing Oil Exporters 1/
Other Oil Exporters 2/
Net Oil Importers 3/
Of which:
Countries not producing oil 4/
3.5
(2.5)
15.2
(8.6)
4.4
-1.9
(3.7)
(-1.8)
18.7
-7.0
(10.8)
(-7.0)
-2.2
(-2.1)
-8.3
(-8.6)
-1.7
(-1.6)
-7.0
(-6.4)
Memo items:
Subsidy Rate by Type of Product (in percent)
for Subsidizing Oil Exporters
Gasoline
9
(8)
Diesel
48
(46)
Residential Light Fuel Oil
71
(78)
Source: IMF staff estimates. Based on IFS exchange rates.
1/ Includes Venezuela and Ecuador.
2/ Includes Argentina, Bolivia, Colombia.
3/ Includes Brazil, Chile, Guatemala, Paraguay, Peru.
4/ Includes Costa Rica, El Salvador, Honduras, Nicaragua, Panama, Uruguay.
Table 2. Domestic Petroleum Price Subsidies in Oil Exprting Countries, 1996-2000
(In percent of GDP)
Country
1996
1997
1998
1999
2000
Algeria
Azerbaijan 1/
Ecuador
Indonesia
Iran
Official exchange rate
Weighted average exchange rate 2/
Kazakhstan
Kuwait
Libya
Mexico
Nigeria
Norway
Qatar
Russian Federation
Saudi Arabia
United Arab Emirates
Venezuela
3.0
..
..
..
2.2
..
3.3
..
..
..
1.4
7.7
2.5
..
-1.5
6.5
..
16.6
12.6
..
10.0
13.2
..
5.1
..
..
..
-4.5
3.9
..
7.4
1.9
..
..
..
..
4.3
5.1
5.8
1.0
..
..
3.2
5.0
7.0
1.4
5.9
..
..
..
-0.9
..
-4.8
..
-5.0
..
..
2.9
4.2
12.6
6.8
4.8
0.7
-0.9
2.5
-3.9
2.4
8.0
5.9
1.7
3.7
..
..
..
-0.4
..
-4.5
..
..
..
..
4.9
Average subsidy
Sample size
4.2
7
4.0
11
0.2
6
3.5
15
7.9
5
Memo item:
Average price of crude oil
(U.S$ barrel, U.K. Brent)
20.5
19.1
12.7
17.7
28.3
Source: EIA, IFS, and authors' estimates
1/ Based on separate IMF Staff estimates.
2/ Based on a trade-weighted average of the official and market exchange rates.
Subsidy Estimates: Results

Subsidies are large — averaged 3.5 percent of GDP in
1999 among major oil exporters

Among net oil importers, the average revenue from
petroleum taxation was 2.2 percent of GDP in 1999.

Subsidies positively correlated with world oil price

Residential fuel oil and auto diesel subsidized more
heavily than gasoline
8
Economic Effects of Subsidies
 Efficiency

In the absence of price distortions, most efficient
to set domestic prices equal to the world
wholesale spot price plus marketing and
distribution costs (W+D)

Caveats:
• Need for government revenue may make net
taxation more efficient
• Environmental externalities may argue for further
taxation (in excess of 100 percent, Parry 2001)
9
Economic Effects of Subsidies

On the other hand, large exporters may have some
monopoly power in world oil markets so that the marginal
revenue from exporting is less than the world wholesale
price plus marketing and distribution costs:
R  (W  D  T ) X
  1  X
dR
dW

W 
X  D  T  W 1  
 X
dX
dX

  X , P  W

  D  T

–R = revenue, X = exports, ηX,P = price elasticity of world oil
demand, and XW = total world oil demand
– Marginal revenue<(W+D) if significant share of world
market (X/ XW is high) and η X,P is low
10
Economic Effects of Subsidies
 Thus,
the opportunity cost of not exporting
may be lower than the world price for
OPEC as a whole or its dominant members.
 The
marginal revenue for OPEC can be as
low as 25 percent of the non-market-power
case.
11
Economic Effects of Subsidies
Table 3. Marginal Revenue from Exporting for Largest Oil Exporters
All OPEC
Saudi Arabia
Russian Fed.
Norway
Venezuela, RB
Exports as Share of World
Consumption 1/
32.7
10.8
5.1
4.3
4.0
(η = -.25)
.25
.75
.88
.90
.91
Marginal Revenue 2/
(η = -.50)
.63
.88
.94
.95
.95
(η = -.75)
.75
.92
.96
.97
.97
Sources: Energy Information Administration (various years); and authors’ estimates.
1/ For 1998 (Energy Information Administration 2001).
2/ As fraction of small exporter (no-market-power) case; assumes distribution and marketing costs are
75 percent of wholesale price.
12
Economic Effects of Subsidies
 Efficiency

On balance, the arguments for taxation
outweigh the arguments for subsidization:
• Need for government revenue
• Environmental externalities

However, most major oil exporters subsidize,
resulting in efficiency (deadweight) losses
• Assuming environmental externalities, deadweight
losses amount to 0.5 to 12.4 percent of GDP
• Assuming no environmental externalities,
deadweight losses amount to between 0 and 7
percent of GDP
13
Economic Effects of Subsidies
 Equity
 Poorly targeted means of distributing purchasing
power to the poor—majority of benefits go to the
non-poor
• In Venezuela, the richest 20 percent receive 6½ times
more in subsidies per person than the poorest 33
percent (World Bank 1995)
• In Ecuador, the more expensive energy products
received the highest subsidies, while household
kerosene, used by poor households, was not subsidized
(World Bank 1994)

Pro-rich bias compounded by possible smuggling
and corruption
14
Table 5. Share of Total Expenditures by Quintile: Mexico and Ecuador
(In percent of total spending)
Quintile
Electricity
Gasoline 1/
Gas/ LPG
Kerex
Others 2/
Mexico
1
2
3
4
5
Ecuador
1
2
3
4
5
5.4
10.2
15.9
21.6
46.8
0.9
3.9
9.3
19.4
66.5
6.7
13.8
18.5
25.1
35.9
15.1
18.3
20.2
21.7
24.8
26.1
24.0
20.4
17.4
12.1
9.2
15.3
16.2
26.4
32.9
…
…
…
…
…
15.2
23.2
20.1
22.1
19.4
43.2
20.6
13.8
7.6
14.8
…
…
…
…
…
Sources: UNDP/World Bank (1994) “Ecuador: Energy Pricing, Poverty and Social Mitigation”
Report No. 12831-EC, p. 40; and IMF staff calculations based on the 1996 Mexican Income and
Expenditure Household Survey; INEGI 1996.
1/ Gasoline refers to gasoline, diesel, or gas purchased as fuel for vehicles.
2/ Includes coal, fire wood, heating oil, candles, and others such as paper or cardboard.
15
Economic Effects of Subsidies
 Fiscal
opportunity costs

In subsidizing countries, the average
subsidy is larger than the average fiscal
deficit or total health spending

In Ecuador and Venezuela, the implicit
subsidy exceeds public spending on health
16
Figure 2. Composition of Government Expenditure in Major
Subsidizing Oil Exporting Countries 1/
6
In percent of GDP
5
4
3
2
1
0
Oil subsidies
(13)
Education Spending
(11)
Health Spending
(12)
1/ Unweighted averages for subsidizing oil-exporting countries. Oil subsidies are for 1999; expenditure
data are an average over the period 1997-99. Number of countries in parentheses.
Table 7. Social Indicators and Social Spending in Subsidizing Oil-Exporting and Other Countries, 1997–1999
Education Spending
Illiteracy Infant
Rate
Mortality
Life
(percent of Rate (per
Expectancy
adult 1,000 live
population) births)
Health Spending
Immunization
Under 5 Rate against GDP Per
Mortality
Measles
Capita
(in
(in percent
(in
(in percent
Rate (per (percent of (constant percent
of total
percent
of total
1,000 live children aged 1995 of GDP) expenditure) of GDP) expenditure)
births) less than 12
US$)
months)
Subsidizing Oil-Exporting Countries (13) 1/
69
20
26
36
88
3,259
5.0
15.7
2.0
5.8
Of which:
Algeria
Ecuador
Indonesia
Nigeria
Saudi Arabia
Venezuela, República Bolivariana de
71
69
65
49
72
73
35
9
14
39
25
8
35
29
43
82
19
21
39
37
54
149
26
24
76
87
80
48
92
79
1,534
1,521
1,025
253
6,841
3,451
8.8
3.1
..
1.4
8.8
3.8
28.0
15.6
..
7.6
23.1
19.7
1.5
0.9
0.6
0.7
2.9
1.4
4.9
4.8
3.3
3.8
7.6
7.0
Low-Income Countries (56)
55
41
75
126
67
423
4.0
15.0
2.0
6.8
Lower-Middle-Income Countries (42)
68
16
33
44
86
1,642
4.8
16.6
2.3
7.5
Upper-Middle-Income Countries (26)
71
11
18
26
90
5,306
4.7
14.4
3.3
9.7
Source: World Development Indicators (World Bank, 2001). Expenditure data from IMF Staff Reports.
1/ Numbers in parentheses refer to the number of countries in the different income groups in 1999.
18
Economic Effects of Subsidies
 Cyclicality

For net oil exporters, periods of high world
oil prices tend to be both periods of
economic boom and periods of higher
domestic subsidies

Petroleum subsidies are thus procyclical,
exacerbating the effects of oil price shocks
on economic volatility
19
Successful Reform

Preconditions



Establish social protection mechanisms to compensate
losers (Indonesia); design using PSIA
Use publicity campaigns to inform public of trade-offs
(Egypt)
Timing




Optimal speed of reform depends on trade-off between
fiscal need and adverse social consequences
Existence of good social protection mechanisms allows
faster reform
Political environment
External environment—periods of low world oil prices
may facilitate movement to automatic price mechanisms20
Conclusion

Major oil-exporting countries tend to be large
net subsidizers of petroleum

Subsidization does not appear to be a wise use
of resources

Reform requires careful design to overcome
political opposition
21