Firms Operating under Electricity Constraints
in Developing Countries
Philippe Alby, Jean-Jacques Dethier, and Ste
´
phane Straub
Many developing countries are unable to provide their industrial sectors with reliable
electric power, with the result that many enterprises must contend with an insufficient
and unreliable supply of electricity. Because of these constraints, enterprises often opt
for self-generation of electricity even though it is widely considered a second-best solu-
tion. This paper develops a theoretical model of investment behavior in remedial infra-
structure in the presence of physical constraints. It then illustrates the model’s
predictions using a large cross-country sample of enterprises from the World Bank
Enterprise Survey database. Electricity-intensive sectors in high-outage countries are
characterized by a significantly lower share of small firms. JEL codes: H54, L94, L16
And God said, ‘Let there be light’ and there was light, but the Electricity Board said He would
have to wait until Thursday to be connected.
Spike Milligan
I NTRODUCTION
Growing evidence, both micro- and macroeconomic, suggests that better elec-
tricity infrastructure significantly boosts economic growth and improves a
range of development outcomes.
1
Low levels of infrastructure development and
Philippe Alby is a Lecturer at the Toulouse School of Economics, ARQADE; his email address is
Philippe.Alby@univ-tlse1.fr. Jean-Jacques Dethier (corresponding author) is a Research Manager at the
World Bank, DEC; his email address is jdethier@worldbank.org. Ste
´
phane Straub is Professor at the
Toulouse School of Economics, ARQADE and IDEI; his email address is stephane.straub@univ-tlse1.fr.
Alby and Straub acknowledge financial support from grant RENTSEP ANR-09-BLAN-0325. We are
grateful to our editor, Elisabeth Sadoulet, to three anonymous referees, and to Thorsten Beck, Alexis
Bonnet, Taryn Dinkelman, Marcel Fafchamps, Vivien Foster, Farid Gasmi, Michael Grimm, Jose
´
Luis
Guasch, Jan Willem Gunning, Franck Lecocq, Jenny Ligthart, Lucio Monari, Nicolas Pistolesi, Franc¸ois
Poinas, Jon Stern, Bernard Tenenbaum, and Clemencia Torres de Mastle as well as the participants in
the 2010 International Conference on Infrastructure and Development in Toulouse and the 2011
Development Economics Workshop at Tilburg University for helpful comments. Any mistakes remain
ours. A supplemental appendix to this article is available at http://wber.oxfordjournals.org/.
1. See, for example, Caldero
´
n and Serve
´
n (2003) and Caldero
´
n (2009) for cross-country estimations
and Dinkelmann (2009) and Lipscomb, Mobarak, and Barham (2011) for microeconomic evidence.
THE WORLD BANK ECONOMIC REVIEW, VOL.27,NO.1,pp. 109–132 doi:10.1093/wber/lhs018
Advance Access Publication June 29, 2012
# The Author 2012. Published by Oxford University Press on behalf of the International Bank
for Reconstruction and Development /
THE WORLD BANK. All rights reserved. For permissions,
please e-mail: journals.permissions@oup.com
109
at International Monetary Fund on January 30, 2013http://wber.oxfordjournals.org/Downloaded from
poor quality services can drive up firms’ costs and bias their technological
choices away from energy intensive choices, increasing the overall costs relative
to competitors in other regions.
2
In developing countries, enterprises typically
have difficulty connecting to the public grid
3
or, when they are connected, face
frequent scheduled and unscheduled power cuts. Fluctuations in voltage and
the frequency of power cuts cause material losses and adversely affect manufac-
turing costs and output.
To offset these negative impacts, industrial firms in developing countries
often opt for self-generation of electricity, even though it is widely considered a
second-best solution. Of the 25 sub-Saharan countries reviewed by Foster and
Steinbucks (2009), in-house generation accounts for more than 25 percent of
the installed generating capacity in three countries and for more than 10
percent in nine others. In Nigeria, where 40 percent of the electricity consumed
is produced with generators, firms spend up to 20–30 percent of their initial
investment enhancing the reliability of their electricity supply.
4
Moreover, in
Africa, self-generated electricity is 313 percent more expensive on average than
electricity from the grid. Foster and Steinbucks (2009) conclude that the main
victims are likely to be the existing informal firms and the formal firms that
were not created as a result of the prevailing constraints.
We use a large firm-level dataset to analyze the behavior of firms facing elec-
tricity constraints. Our objective is to understand the conditions under which
they decide to invest in their own generating capacity, how this decision is af-
fected by the abovementioned constraints, and ultimately, what this decision
implies in terms of firm-size distribution. Our original contribution is, first, to
document systematically the effects of electricity deficiencies on the decision to
invest in mitigating technology (i.e., a generator); second, to analyze how the
impact of electricity deficiencies varies across firms and sectors; and, finally, to
show how these deficiencies affect the resulting patterns of industrial structure
across countries and sectors. A theoretical model of firm responses to power
outages allows us to derive precise predictions that are illustrated by the data.
We show that electricity-related constraints have nonlinear effects that vary ac-
cording to the degree of the sector’s reliance on electricity. In sectors that are
naturally more reliant on electricity, a large number of outages implies a
skewed industrial structure with many large firms and fewer small ones.
Related contributions include Lee, Anas, and Oh (1996) and Lee and others
(1996), both of which use data from Nigeria, Indonesia, and Thailand. These
studies conclude that public infrastructure deficiencies have substantial private
costs with the burden falling disproportionately on smaller firms. They also
point out significant differences across the three countries, particularly in terms
of the regulatory environment. Hallward-Driemeier and Stewart (2004) and
2. Eifert, Gelb, and Ramachandran (2008).
3. World Bank (2004); Bartelsman, Haltiwanger, and Scarpetta (2004).
4. Adenikinju (2005).
110 THE WORLD BANK ECONOMIC REVIEW
at International Monetary Fund on January 30, 2013http://wber.oxfordjournals.org/Downloaded from
Dollar, Hallward-Driemeier, and Mengistae (2005) document patterns of
access to infrastructure services by enterprises in developing countries and
show that access varies by type of service and firm size, with electricity often
being the largest barrier to entry and larger firms expressing more concerns
than smaller firms about all services. In Bangladesh, China, Ethiopia, and
Pakistan, improvements in the reliability of the power supply are found to in-
crease total factor productivity in garment manufacturing as well as output
and employment growth rates. Gulyani (1999) documents the impact of elec-
tricity hazards on an Indian car manufacturer and its upstream suppliers,
which have devised an innovative generation and power-sharing system to
solve their power problems. Reinikka and Svensson (2002) analyze a sample of
171 Ugandan firms, some of which responded to poor electricity supply by in-
vesting in generators, and show that this development came at the cost of
reduced overall investment and less productive capital.
The paper is structured as follows. Section 2 presents the enterprise survey
dataset and provides descriptive statistics on the extent of electricity deficien-
cies. Section 3 presents a model of investment by firms with infrastructure con-
straints. Section 4 presents the econometric specifications used to test the
model’s predictions. Section 5 presents the results, section 6 addresses some
limitations of the data, and section 7 describes selected policy implications and
concludes.
D
ATA AND S TYLIZED F ACTS
We use data from enterprise surveys for 87 countries for which data on the
number of power outages are available, covering a total of 46,606 firms over
the period from 2002 to 2006.
5
For 77 countries, data on generators are avail-
able, and for 34 countries, data on the cost of electricity are available.
6
It is important to note that these surveys are limited to formal registered
firms with five or more employees, primarily in the manufacturing and services
sectors. The implications for informal firms and the effect of the constraints
that we discuss here on formal-informal choice are addressed in several places
below.
Table 1 shows the severity of electricity hazards across regions and country
income groups. Column 1 reports the percentage of firms’ managers who cite
electricity as a major or severe constraint on their operations and growth. This
is the case for more than 26 percent of firms located in low-income countries,
with the highest percentage in South Asia (43 percent of firms), followed by
5. See https://www.enterprisesurveys.org. The list of countries can be found in table S.1 in the
supplemental appendix, available at http://wber.oxfordjournals.org/. Unfortunately, it is not possible to
use survey data after 2006 because key questions about power were dropped from the questionnaire.
6. This yields a sample of 62 countries with data on generators and number of power outages and
32 countries with data on generators, number of power outages, and cost of electricity. The full
description of the variables is in table S.2 in the supplemental appendix.
Alby, Dethier, and Straub 111
at International Monetary Fund on January 30, 2013http://wber.oxfordjournals.org/Downloaded from
East Asia and Africa. Columns 2 and 3 report the average number of power
outages suffered by firms and the share of firms that suffered more than
30 outages in the year prior to the survey. The average number of cutoffs from
the public electricity grid faced by firms per year is as high as 132 in South
Asia and 61 in Africa. Furthermore, in these two regions, close to half of all
the firms surveyed experienced more than 30 outages per year. Note that firms
can report multiple outages per day and that these are counted as separate
episodes.
The picture provided by these indicators is consistent across regions and
income groups. Outages are a particular problem in low-income countries,
with peaks of approximately 250 outages per year in Bangladesh and 180 in
Lebanon and the Democratic Republic of Congo. As a result, many firms
invest in a back-up power generator: 31 percent of all firms own or share a
generator, with this number peaking at 62 percent and 37 percent in
South Asia and Africa, respectively. Note that there is no information in the
dataset on the intensive margin of generator use. Firms with installed generator
capacity are typically larger, are slightly older, and report more days without
power from the public grid during the survey year. Conversely, firms that do
not own generators are smaller and are found mostly in environments with
fewer outages. General firm-level summary statistics are available in table S.3
in the supplemental appendix, available at http://wber.oxfordjournals.org/.
TABLE 1. Access to Electricity by Firms across Regions and Country Income
Groups
Region
Percent of firms
mentioning electricity as
major or severe constraint
Average
number of
power outages
Percent of firms
with more than 30
power outages
Generator
ownership
(percent of firms)
Europe/Central
Asia
8.5% 9.72 5.7% 27.5%
Latin America 9.3% 12.44 7.7% 21.2%
East Asia and
Pacific
25.1% 36,49 18.3% 28.7%
Mid. East/
North Africa
21.5% 41.32 22.1% 32.4%
Sub-Saharan
Africa
16.4% 61.12 45.2% 36.6%
South Asia 43.0% 131.74 49.0% 61.7%
Country Income Level
High 4.9% 1.32 0.2% -
Upper-middle 8.3% 13.02 6.2% 28.0%
Lower-middle 14.3% 13.76 9.1% 24.1%
Low 26.4% 64.08 34.1% 42.4%
Average 15.6% 27.57 15.2% 31.1%
Source: Authors’ analysis based on data described in the text.
112
THE WORLD BANK ECONOMIC REVIEW
at International Monetary Fund on January 30, 2013http://wber.oxfordjournals.org/Downloaded from
Table 2 presents estimations, including firm characteristics and country fixed
effects. In regressions without characteristics, the fraction of the overall vari-
ance of generator ownership explained by country fixed effects is 19 percent,
while for outages variance, it varies between 12 percent (among firms with a
generator) and 28 percent (for firms without a generator).
Firm size explains approximately 10 percent of the variance in generator
ownership, whereas ownership and exporting status explain between 1 percent
and 3 percent. In contrast, none of the variance in outages is due to these char-
acteristics. Large firms as well as foreign-owned, exporting, and capital city-
based firms report owning a generator relatively more frequently than other
firms, and all coefficients are highly significant. Regarding outages, the picture
is more complex. In the sample of firms that own generators, large firms
appear to face more outages than smaller ones, whereas the opposite is true for
firms without generators. This pattern is largely the result of selection because
large firms that are more exposed to outages are also more likely to invest in
self-generation, whereas small firms that would benefit from a generator may
not be able to afford one. Moreover, most of these characteristics, particularly
that the characteristic of a capital city location, are not significant. Together
with the fact that these characteristics fail to explain a significant share of the
overall variance, this leads to limited concerns about the potential endogeneity
of outages, which would derive, for example, from strategic decisions by gov-
ernments about how and where to allocate outages or by firms about where to
locate.
T
HE M ODEL
We are primarily interested in the link between firms’ decisions to invest in self-
generation and the quality of the electricity supply (measured by the number of
outages). The model develops a simple moral hazard credit framework, which
shows that this link crucially depends on the firm’s net worth and its depen-
dence on a reliable electric supply. As shown formally below, when losses
related to electricity deficiencies are limited, firms at all levels of assets are able
to operate, and only those above a given size threshold find it profitable to
invest in a generator. This threshold increases in the level of electricity efficien-
cy. However, when such losses are high, firms that do not own a generator
have very low returns and become unable to operate. Thus, the model provides
predictions with respect to the structure of the industry, particularly the fact
that when facing a large number of outages, sectors that are sensitive to the ef-
ficiency of the electric supply should be characterized by a lower proportion of
small formal firms.
We consider a continuous moral hazard investment model, a
`
la Holmstrom
and Tirole (1997). Firms are endowed with assets A, such as cash or productive
assets that they can pledge as collateral. To undertake an investment project of
variable size I, firms intend to borrow an amount I 2 A.
Alby, Dethier, and Straub 113
at International Monetary Fund on January 30, 2013http://wber.oxfordjournals.org/Downloaded from