Instruments of Political Control

• Author: Johannes Urpelainen,Maureen Lackner,Andrew Cheon
• Published date: 01 March 2015
• Date: 01 March 2015
• DOI: http://doi.org/10.1177/0010414014543440
• Subject Matter: Articles

543440CPSXXX10.1177/0010414014543440Comparative Political StudiesCheon et al.
research-article2014
Article
Comparative Political Studies
2015, Vol. 48(3) 370 –402
Instruments of Political
© The Author(s) 2014
Reprints and permissions:
Control: National Oil
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0010414014543440
cps.sagepub.com
Companies, Oil Prices,
and Petroleum Subsidies
Andrew Cheon1, Maureen Lackner1, and
Johannes Urpelainen1
Abstract
Global petroleum subsidies peaked at US$520 billion in the summer of 2008
and reached US$212 billion in 2011, carrying high fiscal and environmental
costs. Why do some countries spend so much money to subsidize petroleum
consumption? Previous studies suggest that oil-rich autocracies lacking
institutional capacity are the main culprits. However, they cannot explain
why oil importers with capable bureaucracies, such as Argentina, Brazil,
and Malaysia, subsidized petroleum products. We argue that governments
in countries with national oil companies (NOCs) use petroleum subsidies
to cushion the effects of increasing oil prices. Empirically, we examine the
relationship between oil prices and domestic gasoline prices in 175 countries,
2002-2009. An NOC halves the effect of oil price increases on the domestic
gasoline price. This effect is strongly associated with the institutional design
of NOCs, as increased autonomy shields them from political interference by
the government.
Keywords
national oil companies, fuel subsidies, energy policy,oil prices
1Columbia University, New York, NY, USA
Corresponding Author:
Andrew Cheon, Department of Political Science, Columbia University, 420 West 118th
Street, 7th Floor, International Affairs Building, New York, NY 10027, USA.
Email: asc2159@columbia.edu

---

Cheon et al.
371
Introduction
Petroleum subsidies present major environmental and economic problems.
According to International Monetary Fund researchers, the global cost of pre-
tax petroleum subsidies peaked at US$520 billion in the summer of 2008,
amounting to 0.7% of the planet’s gross domestic product (Coady et al.,
2010). In more recent data, the global cost stood at US$212 billion in the year
2011 (International Monetary Fund, 2013b). These subsidies increase the rate
of global warming, produce environmental pollution, contribute to fiscal
insolvency, and divert resources from productive public investment.
Petroleum subsidies are an instrument of political control (Hartley &
Medlock, 2008; Overland, 2010; Victor, 2009). Similar to many other eco-
nomic instruments, such as trade tariffs and regulations, as well as other sub-
sidies (Baldwin, 1989; Buchanan & Tullock, 1975; Ehrlich, 2007; Keohane,
Revesz, & Stavins, 1998; Kono, 2006; Romer & Rosenthal, 1978; Shleifer &
Vishny, 1994), petroleum subsidies allow the government to produce targeted
benefits to politically influential groups and constituencies, meaning that the
political-economic perspective may prove helpful for explaining the high
levels and variation in petroleum subsidies. Governments use subsidies to
furnish valuable benefits to their core supporters, such as urban households
with automobiles. Victor (2009) argues that autocratic governments, espe-
cially those with little institutional capacity to implement competent public
policies, are particularly tempted to use petroleum subsidies to secure the
political support of the public (Overland, 2010). Moreover, petroleum subsi-
dies are high among major oil producers, especially in the Middle East and
North Africa (Fattouh & El-Katiri, 2012).
While existing explanations capture a lot of variation in the data, impor-
tant puzzles remain. If oil producers without bureaucratic capacities are
expected to provide petroleum subsidies, why would countries like Argentina
and Brazil, equipped with more bureaucratic capacity than the vast majority
of developing countries, offer generous petroleum subsidies (International
Monetary Fund, 2013a)? In Argentina, the price of oil is 70% of its interna-
tional reference price. Gas and electricity prices have fared worse at only
20% and 30%, respectively. From 2002 to 2009, oil and gas subsidies cost
Argentina US$55 billion (Fernández-Lamela, 2011). In Brazil of the mid-
1990s, the national oil company (NOC) Petrobras was in enormous debt and
the government was forced to transfer US$5.8 billion to the company as com-
pensation for selling oil below market price (International Monetary Fund,
2013a). These cases are puzzling because significant bureaucratic capacity
should allow for the replacement of inefficient and inequitable petroleum
subsidies with superior policy instruments.

---

372
Comparative Political Studies 48(3)
We propose that NOCs play a facilitating role in the allocation of petro-
leum subsidies. They can be defined by their commitment to some “national
purpose,” meaning political and economic goals other than maximizing the
firm’s profits (Tordo, Tracy, & Arfaa, 2011). NOCs are substantively impor-
tant actors not only because they serve various political-economic functions
for the government but also because they control approximately 90% of the
world’s oil reserves and 75% of production, as well as many of the major oil
infrastructure systems (Tordo et al., 2011). 13 of 14 former and current mem-
bers of the Organization for Petroleum Exporting Countries (OPECs hereaf-
ter) have an NOC, and so do 64 countries in our data set.
An exogenous increase in oil prices helps us capture the role of NOCs in
facilitating the distribution of petroleum subsidies. As petroleum subsidies
are costly, governments face a dilemma in trying to cushion the effects of ris-
ing oil prices through subsidization of petroleum. As oil prices increase, the
political demand for petroleum subsidies grows while economic performance
decreases. If the government’s subsidy policy is ad valorem, the cost of a
given subsidy automatically increases. However, if the government has an
NOC, these problems are both mitigated and hidden. The NOC benefits from
increased oil prices because the value of domestic and foreign reserves
increases. Moreover, the government can directly instruct the NOC to sell oil
at a lower cost to domestic consumers. A form of “optimal obfuscation”
(Kono, 2006), this practice allows the government to hide the subsidy from
the budget (Aharoni & Ascher, 1998), in addition to avoiding legislative
debate. Empirically, we expect domestic gasoline prices to increase less with
international prices in the presence of an NOC.
We expect gasoline prices in countries with an NOC to increase less with
oil prices than in countries without an NOC. We use data on national gasoline
prices in 175 countries for the years 2002-2009 (Coady et al., 2010). Gasoline
prices are determined by oil prices, the cost of transportation, and subsidies/
taxes. As international markets are largely integrated and transportation costs
low, prices across different countries should converge to the world reference
price in the absence of government intervention (Koplow, 2009). The pres-
ence of an NOC would thus influence gasoline prices through subsidies and
taxes. We focus on gasoline because it is the most important petroleum prod-
uct, and the time period is ideal for testing the argument because there is
significant variation in oil prices.
An increase in the international oil price produces a corresponding increase
in domestic gasoline prices that is half as large in the presence of an NOC
than otherwise. Suggesting a large substantive effect, the result is robust to
variations in model specification. It holds even when we specifically account
for OPEC and non-OPEC oil producers, suggesting that oil wealth itself is

---

Cheon et al.
373
not behind the finding that NOCs play a major role in raising petroleum sub-
sidies in times of high oil prices. The finding holds if we use 20-year lags to
instrument for the presence of an NOC, and qualitative case studies confirm
the pattern.
Using the original NOC data we collected, we also explore the role of
institutional design. One of the key findings from this analysis is that as the
government’s institutional hold over the NOC weakens, the NOC becomes
less useful to the government as an instrument of control over the economy.
We explore, in particular, the importance of the government’s ability to
appoint the chairman of the NOC and directly control the NOC’s budget,
finding that independent chairmen and budgetary autonomy reduce the
NOC’s role as an instrument of political control. These findings are also
important for policy in that they show how institutional design can insulate
the NOC from governmental intervention.
This study makes four key contributions. First, while the existing litera-
ture recognizes that NOCs are the main instrument of government control
over the energy sector and they often play a role in the allocation of fuel
subsidies (Hartley & Medlock, 2008; Khan, 1987; Losman, 2010; Luong &
Weinthal, 2010; Olorunfemi, 1991; Philip, 1982; Ross, 2012; Victor, Hults, &
Thurber, 2012), systematic studies of NOC involvement in fuel subsidies do
not exist. For instance, the existing literature draws heavily from the experi-
ence of oil producers, which may not apply to non-producers. NOCs also
exist in oil-dependent economies, and we show that they have a role to play
in their fuel subsidies (Jaffe & Soligo, 2010). Without contesting the role of
other factors, such as authoritarian institutions and the lack of...