A partnership with academia

Building knowledge for trade and development

Vi Digital Library - Text Preview

Infrastructure and Employment Creation in the Middle East and North Africa

Case study by Antonio Estache, Elena Ianchovichina, Robert Bacon,Ilhem Salamon, 2013

Download original document (English)

The study discusses how infrastructure investments can help to stimulate employment creation in the immediate future while building foundations for sustainable growth and job creation in the MENA region.

Pu
bl


ic
Di


sc
lo


su
re


A
ut


ho
riz


ed
Pu


bl
ic


Di
sc


lo
su


re
A


ut
ho


riz
ed


Pu
bl


ic
Di


sc
lo


su
re


A
ut


ho
riz


ed
Pu


bl
ic


Di
sc


lo
su


re
A


ut
ho


riz
ed




Infrastructure and Employment Creation in
the Middle East and North Africa






Infrastructure and
Employment Creation in the
Middle East and North Africa
Antonio Estache
Elena Ianchovichina
Robert Bacon
Ilhem Salamon




© 2013 International Bank for Reconstruction and Development / The World Bank
1818 H Street NW
Washington DC 20433
Telephone: 202-473-1000
Internet: www.worldbank.org


Some rights reserved


1 2 3 4 16 15 14 13


This work is a product of the staff of The World Bank with external contributions. Note that The
World Bank does not necessarily own each component of the content included in the work. The
World Bank therefore does not warrant that the use of the content contained in the work will not
infringe on the rights of third parties. The risk of claims resulting from such infringement rests solely
with you.


The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the
views of The World Bank, its Board of Executive Directors, or the governments they represent. The
World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors,
denominations, and other information shown on any map in this work do not imply any judgment
on the part of The World Bank concerning the legal status of any territory or the endorsement or
acceptance of such boundaries.


Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges
and immunities of The World Bank, all of which are specifically reserved.


Rights and Permissions


This work is available under the Creative Commons Attribution 3.0 Unported license (CC BY 3.0)
http://creativecommons.org/licenses/by/3.0. Under the Creative Commons Attribution license, you
are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under
the following conditions:


Attribution—Please cite the work as follows: Estache, Antonio, Elena Ianchovichina, Robert Bacon,
and Ilhem Salamon. 2013. Infrastructure and Employment Creation in the Middle East and North
Africa. Washington, DC: World Bank. doi:10.1596/978-0-8213-9665-0. License: Creative Commons
Attribution CC BY 3.0


Translations—If you create a translation of this work, please add the following disclaimer along with
the attribution: This translation was not created by The World Bank and should not be considered an
official World Bank translation. The World Bank shall not be liable for any content or error in this
translation.


All queries on rights and licenses should be addressed to the Office of the Publisher, The World Bank,
1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank
.org.


ISBN (paper): 978-0-8213-9665-0
ISBN (electronic): 978-0-8213-9666-7
DOI: 10.1596/978-0-8213-9665-0


Cover photo: Workers on scaffolding in Dubai © Tom Horton/Further To Fly Photography
Cover design: Naylor Design


Library of Congress Cataloging-in-Publication Data has been requested.




v


Contents


Foreword xi
Acknowledgments xiii
About the Authors xv
Abbreviations xvii
Overview xix


Chapter 1 Introduction 1
Background Information 3
Scope and Structure of the Study 5
Definitions and Key Concepts 6
References 8


Chapter 2 The State of Employment and
Infrastructure and Future Needs 9
Employment Challenges 9
Infrastructure Endowments and Future Needs 11
Potential of Infrastructure Investment in


Boosting Employment 19
Annex 2A Econometric Models for


Infrastructure Needs 23




vi   Contents


Annex 2B Data Sources and Descriptions Used
for Model of Investment Requirements 24


Annex 2C Data Imputations 25
Notes 25
References 27


Chapter 3 Short-Run Employment Effects
of Infrastructure Investment 29
Techniques for Estimating the Cost of a Job and


the Employment Generated by Investment in
Infrastructure 29


Hybrid Approaches to Estimating the
Short-Term Employment Effects
of Infrastructure Investment 32


Estimating the Cost of Creating Jobs in
Oil Importing MENA Countries 35


Alternative Approaches to Estimating the
Short-Term Employment Effects of
Infrastructure Spending 45


Implications of Using Labor-Intensive
Technologies in the Maintenance of
Unpaved Roads 49


Annex 3A Constructing Hybrid Estimates of
Employment Linked to Investment 51


Annex 3B Estimated Shares of Inputs in
Different Types of Infrastructure 55


Annex 3C Potential for Job Creation in the
Three Groups of MENA Countries 55


Notes 57
References 58


Chapter 4 Long-Term Employment Effects through
the Growth Channel 61
Output Elasticity with Respect to Infrastructure 62
Employment Elasticity with Respect to


Economic Output 63
Employment Elasticity with Respect to


Infrastructure 64
Long-Run Employment Response to


Infrastructure Investment 66




Contents   vii


Notes 67
References 67


Chapter 5 Policy Implications 69
Subsidized Employment Programs and


Job Creation 70
Types of Training for Lasting Job Creation 71
Minimizing the Cost of Job Creation


Targeting 73
Is Subsidizing Job Creation a Sustainable Policy? 73
What Are the Net Fiscal Costs and Benefits of


Job Creation Programs? 76
Concluding Remarks 77
Notes 81
References 82


Boxes
3.1 Pros and Cons of Input-Output Table Use for


Generating Employment Estimates 33
3A.1 Calculation of the Cost of a Type II Job Using a


Hybrid Approach 54


Figures
O.1 Infrastructure Needs and Financing xxi
O.2 Cost of a Direct Job in Roads and Bridge Construction


Relative to Other Sectors in 2009 xxiii
O.3 Shares of Unemployed by Education Level in


Selected MENA Economies xxvi
1.1 Public Gross Fixed Capital Formation 3
1.2 Sectors’ Contribution to Annual Employment


Growth in the 2000s 4
1.3 Fiscal Space Indicators 5
2.1 Labor Force Levels in MENA, 2009 11
2.2 Composition of Infrastructure Expenditure Needs by


Group of Countries 17
2.3 Shares of Infrastructure and Construction Jobs in


Total Employment in MENA 22
3.1 Hourly Wages as a Function of the Share of Labor


Inputs in Total Costs 47




viii   Contents


Tables
O.1 Infrastructure-Related Short-Term Job Creation xxiii
1.1 MENA Classification 6
2.1 Employment and the Size of the Labor Force in MENA 10
2.2 Infrastructure Endowments in the Developing World 12
2.3 Infrastructure Endowments in MENA by Country Grouping 12
2.4 Unit Costs of Infrastructure by Sector 14
2.5 Roads Maintenance and Rehabilitation Program 15
2.6 Annual Expenditure Needs for Infrastructure in


the MENA Region 16
2.7 Annual Infrastructure Investment and Maintenance Needs


in MENA by Type of Investment as Percent of GDP 18
2.8 Expenditure Needs for Water and Sanitation in


Urban and Rural Areas 18
2.9 Access Shortfall Compared to MDG Linear Path


Achievement 19
2.10 Employment Shares of Infrastructure and


Construction Sectors 20
2.11 Infrastructure Jobs by Sector in MENA 21
2C.1 Imputation of Average Investment as Percent of GDP


When Data Were Not Available 25
3.1 Regression of Semiskilled Hourly Construction Wage on


GDP per Capita 34
3.2 Construction Sector Hourly Wages in the


Arab Republic of Egypt, January 2009 35
3.3 Sector Coverage Provided by Various


Input-Output Tables 36
3.4 Cost of Creating a Job in Selected Infrastructure


Sectors in the Arab Republic of Egypt, 2009 37
3.5 Cost of Creating Infrastructure-Related Jobs by Country 39
3.6 Estimated Costs of a Type II Job in Six MENA OICs, 2009 41
3.7 Estimated Costs of a Direct Job in Six MENA OICs, 2009 42
3.8 Number of Type II Jobs Generated per US$1 Billion of


Spending, 2009 43
3.9 Shares of Total Investment Needs by Sector for OIC 44
3.10 Number of Type II Jobs Created by a US$1 Billion


Portfolio of Infrastructure Spending 44
3.11 Estimated Hourly Wages in Infrastructure Works, 2010 46
3.12 Effect of US$1 Billion of Infrastructure Investment on


Job Creation in MENA 48




Contents   ix


3.13 Estimated Potential Job Creation in Response to
Meeting Infrastructure Needs in MENA 49


3.14 Road Maintenance and Rehabilitation Program by
Type of Technology 50


3.15 Investment Needs for Unpaved Road Maintenance by
Type of Technology 51


4.1 Studies Providing Estimates of the Output Elasticity
with Respect to Infrastructure 63


4.2 Employment Elasticities with Respect to GDP in
MENA, 2009 64


4.3 Lower and Upper Bounds for the Employment Elasticity
with Respect to Infrastructure 65


4.4 Employment Response to Infrastructure Investment
Resulting in a Percentage Point Additional Growth 67






xi


Political transitions have lifted expectations in the Middle East and North
Africa (MENA) region for rapid improvements in the population’s well-
being. The issue of jobs is central to well-being, but progress to date has
been insufficient to address the needs of the fast-growing labor force. The
transition period has been especially challenging. Since the beginning of
the Arab Spring, unemployment rates have soared in many transition
economies while private investment and economic growth have declined
and government finances have deteriorated. Economic weakness in key
European markets and the slow global economic recovery have added to
domestic pressures.


Achieving tangible results relatively quickly on the jobs agenda has,
therefore, become imperative in the MENA region. An important ques-
tion facing governments in the region is how to stimulate employment
creation in the immediate future while building foundations for sustain-
able growth and job creation. Infrastructure and Employment Creation in
the Middle East and North Africa discusses how infrastructure investments
can be part of the solution to this problem. The authors find the potential
of infrastructure investments to create jobs to be significant, with some
countries and some sectors having a much greater potential than others.


The report outlines how policy makers can make infrastructure invest-
ment an effective instrument of job creation in the face of fiscal pressures


Foreword




xii   Foreword


to reduce spending. Infrastructure projects will need to be prioritized
based on a country’s employment and infrastructure needs to boost the
short-term job creation impact of public investment programs, while
building the foundation for long-run growth. Importantly, strengthening
governance and the efficiency of investment spending will be critical for
delivering results through infrastructure investments.


Infrastructure and Employment Creation in the Middle East and North
Africa offers new and compelling evidence on the potential for infrastruc-
ture investment to create jobs, while meeting country needs for better
transport, communications, and housing. The authors show that, in the
short-run, every US$1 billion invested in infrastructure has the potential
to generate, on average, 110,000 infrastructure-related jobs in the oil
importing countries; 26,000 jobs in the Gulf Cooperation Council econo-
mies; and 49,000 jobs in the developing oil exporting countries. With
estimated annual infrastructure needs of about US$106 billion, the
region could generate 2.5 million jobs by meeting these needs. However,
these jobs would be lost if countries instead decide to trim their public
investment rates going forward.


Despite these gains, infrastructure alone cannot resolve the substantial
employment challenges in the region. As a complement to labor-intensive
public works, subsidized employment programs combined with training
and counseling can also be used to create jobs. Countries will also need
to take action on a broader set of reforms aimed at generating a more
dynamic private sector, by improving business regulations, promoting
access to finance, and building more transparent and accountable institu-
tions. While such reforms take time to demonstrate results, infrastructure
investment can help maintain confidence, providing immediate gains on
jobs and tangible improvements in the environment.


Infrastructure investment can provide a quick response, helping to
build confidence by creating jobs during the transition period, while put-
ting in place the roads, housing, utilities, and communications platforms
necessary for long-run growth. We hope that the analysis and data pre-
sented on alternative types of investment across the region will be useful
in helping policy makers to implement the infrastructure programs that
best fit their needs.


Caroline Freund
Chief Economist


Middle East and North Africa Region




xiii


This study was prepared by a team led by Elena Ianchovichina (Lead
Economist, MNACE, and principal co-author) and Ilhem Salamon
(Senior Energy Economist, MNSEG, and principal co-author), and com-
prised Robert Bacon (principal co-author), Antonio Estache (principal
co-author), Tito Yepes, Grégoire Garsous, Renaud Foucart, and Caroline
Bahnson, all consultants for the World Bank. The study also builds on the
outcomes of background papers developed by Robert Bacon, Antonio
Estache, Renaud Foucart, and Grégoire Garsous, as well as Hayat Taleb
Al-Harazi (Operation Analyst), Josef Loening (Economist), Deepali
Tewari (Senior Municipal Development Specialist), Cecilia Maria Paradi-
Guilford (ET Consultant), Caroline Van Den Berg (Lead Water and
Sanitation Specialist), and Vincent Vesin (Transport Specialist).


The study is a joint effort of the Chief Economist Office and
Sustainable Development Unit of the World Bank’s Middle East and
North Africa (MENA) region and was prepared under the guidance of
Caroline Freund (Chief Economist, MENA Region), Laszlo Lovei (Sector
Director, MNSSD), and Jonathan Walters (Regional Strategy and Programs
Director, MNARS). The study also benefited from the comments and
guidance of the three peer reviewers: Moustafa Baher El-Hifnawy (Lead
Transport Specialist), Jordan Schwartz (Lead Economist), and Rebekka
Grun (Senior Economist).


Acknowledgments






xv


Antonio Estache is professor of economics at Université Libre de
Bruxelles in Belgium and a research fellow at the European Center for
Advanced Research in Economics and Statistics (ECARES), Brussels, and
the Center for Economic Policy Research (CEPR) in London. He has also
served as chief economist of the World Bank’s Sustainable Development
Vice Presidency and as senior economic advisor for the Poverty Reduction
and Economic Management Vice Presidency.


Elena Ianchovichina is lead economist at the Chief Economist Office of
the World Bank’s Middle East and North Africa region. Prior to this, she
managed the program on inclusive growth in the Economic Policy and
Debt Department of the World Bank. She has also served in the World
Bank’s Research Department and East Asia and Pacific region.


Robert Bacon is a consultant at the World Bank Group and an expert on
infrastructure and energy issues. He has extensive experience on these
issues in different country contexts.


Ilhem Salamon is a senior energy economist in the Energy Sector Group
of the World Bank’s Middle East and North Africa region. She also has
experience working on sustainable development issues in Africa.


About the Authors






xvii


CIM Construction, installation, and manufacture
FTE Full-time equivalents
GCC Gulf Cooperation Council
GDP Gross domestic product
ICT Information and communication technology
ILO International Labour Organization
IMF International Monetary Fund
IO Input-output
ISIC International standard industrial classification
LAC Latin America and Caribbean
LE Egyptian Pound
MDG Millennium Development Goal
MENA Middle East and North Africa
O&M Operations and maintenance
OEC Developing oil exporting country
OIC Oil importing country
TEU Twenty-foot equivalent units (containers)
WDI World Development Indicators


Abbreviations






xix


General Context


Lack of job opportunities, especially for young people, is a well-known,
major issue in the Middle East and North Africa (MENA). The region’s
labor force has been growing at a rapid pace—a consequence of relatively
high population growth over the years and rising female labor force par-
ticipation, but job creation has been lagging. This study assesses the
potential for job creation through infrastructure investment in the
MENA region. The need to achieve tangible employment results rela-
tively quickly has become urgent in the context of the Arab Spring
events. Moreover, heightened regional and global uncertainty has tempo-
rarily restrained private investment—the traditional source of new jobs in
expanding economies.


Effectively directed and fostered, infrastructure investment has a deep
and far-reaching impact on economic development. Infrastructure proj-
ects can serve as a potential source of immediate jobs and can boost long-
term growth and employment. They can also help meet social goals.
Improved provision of high-quality basic infrastructure services, such as
hospitals, schools, and water supply and sanitation, raises living standards
and improves employability of populations and prospects for inclusive
growth.


Overview




xx   Overview


MENA countries already have experience in making huge infrastruc-
ture investments. The region has indeed been investing in infrastructure
over the years. Both in the 1990s and 2000s, public investment spending
in MENA was higher than in most developing regions, largely because of
robust spending in the oil exporting countries, which benefited from ris-
ing fuel prices. Spending on infrastructure boosted employment in the
construction sector, which was a major source of job growth in the 2000s
relative to other sectors and countries. The study shows that maintaining
and spreading the momentum in infrastructure investment will be impor-
tant to support growth and job creation. To do so, policy makers will have
to recognize that there are large differences in the initial conditions across
the region in terms of starting stocks, needs, fiscal commitments, and
potential for job creation.


Variations in the Status of Infrastructure Development
Although infrastructure investment in the region overall has been strong,
there is wide variation across countries in the quality and quantity of
infrastructure. The high-income Gulf Cooperation Council (GCC)
group has the best infrastructure endowments and services in the region,
reflecting advanced stage of development and commitment to infra-
structure investments financed by oil revenues. However, infrastructure
deficiencies in developing MENA remain a concern. Public investment
spending has been particularly weak in the oil importing countries
(OICs), which have much more limited fiscal space than the oil export-
ing countries.


While public investment rates increased in the oil exporting countries
in the 2000s relative to the 1990s, the opposite happened in the OICs.
Recent growth in public-private partnerships was beginning to close the
gap in some OICs, but the economic consequences of the Arab uprisings,
combined with economic difficulties in Europe, have strained fiscal bud-
gets in developing MENA and reduced private investment, with possible
negative consequences for infrastructure spending.


The complexity of identifying infrastructure needs stems not only
from differences in the quality and quantity of infrastructure endow-
ments and services across countries but also from differences in needs
within countries and sectors. Moreover, under a business as usual sce-
nario, the gaps are likely to magnify as demand for infrastructure grows
with population and income growth, and countries face challenges
related to water and energy conservation, efficiency, and climate change.




Overview   xxi


Variations in Infrastructure Needs
The overall needs are quite large. This study estimates MENA’s infrastruc-
ture investment and maintenance needs through 2020 at about US$106
billion per year or 6.9 percent of the annual regional gross domestic prod-
uct (GDP). The estimated differences in needs across subregions are just
as impressive. Developing oil exporting countries (OECs) are expected to
commit almost 11 percent of their GDP annually (US$48 billion) on
improving and maintaining their national infrastructure endowments,
whereas the OICs and the GCC oil exporters will need approximately 6
and 5 percent of their GDP, respectively, to ensure enough infrastructure
to meet their growth and poverty reduction targets (figure O.1).


Investment and rehabilitation needs are likely to be especially high in
the electricity and transport sectors, particularly roads. Electricity and trans-
port are each estimated to account for about 43 percent of total infrastruc-
ture needs in MENA, followed by information and communication
technology (9 percent) and water and sanitation (5 percent). Fulfilling the
electricity need alone would require approximately 3 percent of the annual
regional GDP, or US$46 billion, of which US$10 billion will be spent in
OICs and around US$36 billion in OECs. During the next decade, devel-
oping OICs in MENA will need to spend about US$86 billion on
upgrading their transport networks, whereas the developing and GCC oil
exporters will need US$225 billion and US$145 billion, respectively.
Rehabilitation needs are expected to account for slightly more than half
of the total infrastructure needs.


Figure O.1 Infrastructure Needs and Financing


0


2


4


6


8


10


12


14


MENA


A
n


n
u


al
, p


er
ce


n
ta


g
e


o
f G


D
P


Oil importers Other oil exporters GCC


Gap


Infrastructure needs, average next decade
Total investment spending, average 2000


Sources: World Bank and World Bank Private Participation Infrastructure Database.
Note: GCC = Gulf Cooperation Council; GDP = gross domestic product; MENA = Middle East and North Africa.




xxii   Overview


Variations in Commitments to Meet the Infrastructure Needs
While oil exporters will be able to meet their national infrastructure
needs if they maintain investment spending at rates prevailing in the
2000s, oil importers will fall short. Since the vast majority of funding for
infrastructure comes from public budgets, it will be critical to protect
public investment budgets and try to increase resources going to the sec-
tor in the case of oil importers. Doing so will be a smart choice for gov-
ernments looking to create jobs and growth. The fiscal challenge will be
the toughest for the poorest countries of the region since they are the
least likely to be able to attract private financing for infrastructure needed
to meet the needs of populations.


Variations in Infrastructure’s Employment Potential
MENA’s infrastructure sectors, including construction and infrastructure
services, employ close to one-fifth of the regional workforce or 18.2 mil-
lion people; of these, 10.7 million workers are employed in the construc-
tion sector, whereas the remaining 7.5 million provide infrastructure
services. Within infrastructure services, the transport and communication
sectors are the biggest employers, representing jointly about 7 percent of
the total employment, with energy and water sectors accounting for
approximately 1 percent. These aggregate numbers hide significant varia-
tions across countries, as the Islamic Republic of Iran, for example,
employs more than 40 percent of the country’s workforce in the con-
struction and infrastructure sectors, whereas the Arab Republic of Egypt
and the Republic of Yemen employ just around 10 percent.


In addition to being a large employer, the infrastructure sector has the
potential to contribute to employment creation in MENA, although it
alone will not resolve the region’s unemployment problem. In the short
run, every US$1 billion invested in infrastructure has the potential of gen-
erating, on average, around 110,000 infrastructure-related jobs in the
OICs, close to 49,000 jobs in the OECs, and approximately 26,000 jobs in
the GCC economies (table O.1). The region could therefore generate 2.5
million direct, indirect, and induced infrastructure-related jobs just by
meeting estimated annual investment needs, but the potential varies
greatly across countries, and these jobs account for less than 2 percent of
the labor force in the region. Put differently, these jobs would never mate-
rialize if countries decide to trim their public investment rates going for-
ward. Infrastructure investments could provide a quick response and be
part of the solution to the unemployment challenge, but infrastructure
alone will not resolve this problem.




Overview   xxiii


Because of per capita income differences, spending of US$1 billion
generates more than six times as many jobs in a sector in low-income
Djibouti than in upper middle-income Lebanon, but the latter would
find it easier to finance investment expenditure. Spending on construc-
tion of roads and bridges would generate more jobs as the same amount
of spending in any other infrastructure sector. This is because the cost of
an infrastructure job in the roads and bridge construction sector is less
than one-fifth of the cost of a job in the electricity-generating sector, and
slightly less than one-tenth of the cost of a job in the transport and com-
munication services sector (figure O.2).


Table O.1 Infrastructure-Related Short-Term Job Creation


Infrastructure
needs


(US$billions)


Direct
jobs/


US$billion
Total jobsa/
US$billion


Labor force
(thousands)


in 2009


Direct jobs as
a share of the


labor force
(percent)


Total jobs as a
share of the
labor force
(percent)


GCC 15.8 20,859 26,194 16,387 2.01 2.53
OIC 10.3 86,566 109,236 61,598 1.45 1.83
OEC 20.7 39,454 48,573 52,884 1.54 1.90
Total 46.8 2,037,900b 2,544,457b 130,869 1.56 1.94


Source: World Bank data.
Note: GCC = Gulf Cooperation Council; OEC = developing oil exporting country; OIC = oil importing country.
a. Total jobs include direct, indirect, and induced jobs created per US$1 billion in the short run.
b. The estimate of total direct jobs in the last row of the table refers to the jobs created by meeting annual
infrastructure needs. This estimate is obtained by multiplying the estimated infrastructure needs for a particular
group with the corresponding direct jobs estimated per US$1 billion, and then summing up across groups.


Figure O.2 Cost of a Direct Job in Roads and Bridge Construction Relative
to Other Sectors in 2009


35


30


25


20


15


10


5


0
Water and


sewage
Electricity
stations


Other
construction


Buildings Electricity
services


Transport and
communication


services


40


45


Pe
rc


en
t


Source: World Bank data.




xxiv   Overview


Sectors also differ in their propensity to generate indirect jobs. It
depends on the extent to which the sector requires inputs from other
sectors to produce its output. In Egypt, the ratio of all jobs to the number
of direct jobs was as low as 1.09 for construction in roads and bridges,
whereas it was 1.82 for transport and communication. This indicates that
when investment decisions are made with the objective of creating jobs,
consideration should be given to both direct and indirect employment
effects, as well as the type of skills required to implement projects.


The long-term employment effect of infrastructure investment could
be significant. The study finds that the employment response induced by
infrastructure investment resulting in 1 percentage point additional
growth is expected to be 9 million additional jobs in the course of
10 years in MENA or a little less than 1 million jobs per year. Such a
response is significant as it accounts for approximately 30 percent of the
jobs created in the region during the 2000s. Had these jobs been created
during the last decade, the unemployment rate would be substantially
lower than the 10 percent registered in 2009.


The infrastructure investment required to boost growth by a per-
centage point would vary by country. The lower the growth elasticity
with respect to infrastructure, the higher the required increase in the
stock of infrastructure. For example, the lower bound of the elasticity
suggests that an increase of 8.7 percent in the stock of infrastructure is
required to add a percentage point to growth in the MENA region. This
is the more likely scenario in high-income MENA, comprising the GCC
economies and some upper middle-income MENA countries, whereas
in the more developed countries the likely growth impact of an addi-
tional unit of infrastructure investment tends to be smaller. With the
upper bound elasticity, the required increase in infrastructure stock is
just 3.1 percent.


A switch to labor-intensive technology could enhance the employ-
ment creation effect of infrastructure investment, and it may also reduce
overall costs. The study discusses the possibility of doing so in the main-
tenance of unpaved roads and finds that the use of labor-intensive tech-
nology reduces investment needs in the region by 0.3 percent of GDP.
But, solely focusing on costs is probably not the best criterion when
considering labor-intensive technologies. The cost structure of labor-
intensive infrastructure provision is different from equipment-intensive
alternatives, as it includes components like training or development of
institutional capacity. Direct comparisons of labor versus nonlabor costs
can therefore be misleading.




Overview   xxv


Policy Implications


Infrastructure investment has the potential to create jobs quickly, while
providing a foundation for future growth. This is especially important in
the OICs, where the infrastructure gap is the greatest and employment
needs are growing. However, it is also likely to be most difficult in these
countries because of strained finances. Going forward, government deci-
sions on what types of spending to expand and what to downsize in order
to achieve balanced budgets will have important implications for jobs. In
designing country-specific solutions, governments will have to take on
predictable challenges: the governance of job creation, the proper target-
ing and fiscal cost assessment of subsidies needed to create jobs, the
design and fiscal costs of the (re)training programs needed, and the
expectations on the job creation effects of infrastructure.


Governance Challenge
Prudent infrastructure development will be critical for short- and long-
term growth and job creation because the greatest risk to using infrastruc-
ture as part of an employment and growth strategy in MENA countries
is poor governance. Not all jobs are equal in terms of skills, and not all
infrastructure investments are equal in terms of ability to create jobs for
different skills. This means that investments in infrastructure will need to
be prioritized based on the employment and infrastructure needs and
opportunities in the country. For example, road and bridge construction
projects will have a direct impact on creation of relatively low-skilled
jobs. These types of projects will be especially effective in addressing job-
related concerns in countries where there is a large pool of relatively
unskilled and unemployed nationals. This is the case in most MENA
countries, where the majority of the unemployed do not have tertiary
education (figure O.3). By contrast, projects in transport and communica-
tion services have large indirect effects and, therefore, the ability to create
a diverse set of jobs for workers with different skill levels. These projects
will appeal to policy makers in countries where the unemployed have the
ability to acquire specialized skills relatively quickly.


Subsidy Targeting Challenge
Public works and different types of subsidized employment programs
have been used widely to make it easier for people who cannot find
unsubsidized jobs to find employment and acquire on-the-job skills.
These programs are necessary, for instance, to address structural issues,




xxvi   Overview


which will not be addressed through market forces alone as economies
grow bigger. Subsidies to job creation in infrastructure and construction
will have to be designed to make the most of employment opportunities
for low-skilled workers. The design of the targeting will also have to
address the pressing nature of the need to create jobs. Indeed, boosting
short-term job creation in MENA is desirable, particularly in the context
of the recent political developments. But subsidized employment pro-
grams are costly and should be designed to ensure that there is a positive
spillover to long-run employment and employability.


Employment Subsidies Costing and Financing Challenge
The net costs of subsidizing job creation are difficult to estimate, although
the temporary nature of the subsidies, which last only during the invest-
ment phase of an infrastructure project, minimizes any potential losses.
In addition to the direct fiscal costs of providing the subsidies and any
associated training and program management, there are less obvious costs
in the form of deadweight loss, substitution, and displacement effects.
The costs would also be overestimated if the induced formalization of the
labor market and hence the potential revenue from labor taxes are
ignored, and underestimated if this formalization leads quickly to added
expenses in unemployment benefits and other indirect related costs.
There is also the opportunity cost of how the funds are spent. In an
economy with poor institutional quality and high levels of rent-seeking


Figure O.3 Shares of Unemployed by Education Level in Selected MENA Economies


100


90


80


70


60


50


40


30


20


10


0


Pe
rc


en
t


West Bank
and
Gaza


Iraq Jordan Tunisia Morocco Egypt,
Arab
Rep.


Nationals
of United


Arab Emirates


Tertiary Secondary Primary and below


Source: World Bank data.
Note: MENA = Middle East and North Africa.




Overview   xxvii


behavior, public spending on infrastructure could lead to projects with
low value added and cost overruns. Thus, good governance is a key com-
plement to infrastructure spending.


Training Challenge
Experience shows that the long-term payoffs of employment subsidies
can be achieved only if subsidized employment programs are combined
with training and counseling. Therefore, the design of these programs
should be given as much attention as the design of the subsidized
employment programs. Specific training should be considered only if
there is market demand for these qualifications or if there is a need to buy
time in a labor market restructuring transition. Often, general training
supporting labor market flexibility will be sufficient and more efficient in
increasing productivity than specialized training.


Challenge of Managing Expectations
The study shows that infrastructure investments could provide a rela-
tively quick, short-term response to MENA’s unemployment challenge.
As such, it is part of the solution, but infrastructure alone will not resolve
the problem. Infrastructure and construction jobs represent less than
20 percent of the jobs in most countries of the region. Even a dramatic
increase in labor-intensive infrastructure investments and maintenance
would not be able to address the very large unemployment rate of the
region. Countries should press on with reforms that improve the business
environment, especially business regulations and governance. The impor-
tance of a sound regulatory environment and good governance for inclu-
sive growth has been underscored in numerous studies. This study
focused on estimating the employment impact of infrastructure invest-
ment in MENA. In the future, more work needs to be done to assess the
impact of infrastructure investment on different types of labor, for
example, skilled versus unskilled, young versus old, and domestic versus
migrant workers.






1


C H A P T E R 1


Introduction


The state of national labor markets has always been a concern for govern-
ments and development agencies such as the World Bank. Key labor mar-
ket indicators, such as the rate of unemployment, send signals about the
health of an economy and mirror citizens’ attitudes. Being gainfully
employed is an important aspect of an individual’s well-being both finan-
cially and socially, as “initial failures in finding a job can lead to persistent
joblessness, a loss of interest in further schooling, delayed family forma-
tion, mental distress, and negative manifestations of citizenship” (World
Bank 2007). Importantly, high unemployment tends to increase the risk of
violence, and unemployment and idleness are the most cited reasons for
young people to join gangs and rebel groups (World Bank 2011a).


The Arab Spring events of 2011 brought to the fore concerns about high
unemployment, especially among the youth in the countries of the Middle
East and North Africa (MENA). This region is facing daunting employment
challenges. The unemployment rate in MENA has been higher than in any
other region in the world, and has been especially high among youth (World
Bank 2004). At the same time, the labor force has been growing at a rapid
pace—the consequence of relatively high population growth over the years
and also increasing female labor force participation.


Reforms intended at improving the competitiveness and the invest-
ment climate of MENA economies and strengthening the employability




2 Infrastructure and Employment Creation in the Middle East and North Africa


of workers are key factors in boosting job creation, especially for young
people in MENA. However, the effects of such reforms would only be
felt gradually, whereas the urgency of the employment challenge in the
context of the Arab Spring events calls for policies that have an immedi-
ate impact on job creation. These will be particularly helpful, given the
heightened regional and global uncertainty, which has temporarily
restrained private investment—the traditional source of new jobs in
expanding economies. This report therefore focuses on investment in
infrastructure, which has the potential to create jobs quickly, while pro-
viding a foundation for future growth.


It is well known that, if effectively directed and fostered, infrastructure
investments will have deep and far-reaching impacts on economic and
social development. Infrastructure investments have the potential to
immediately boost employment by creating jobs in construction and infra-
structure services. These investments in turn boost employment in sectors
that either supply inputs to infrastructure projects directly or indirectly, or
supply goods and services to meet the extra demand created by the addi-
tional income of those benefiting directly and indirectly from the infra-
structure spending. The infrastructure built by these investments in turn
increases the need for sustainable employment in operation and mainte-
nance and creates a foundation for growth of enterprises, enabling them
to expand operations and employment. The social payoff of developing
sustainable and integrated basic infrastructure is also significant. Improved
provision of high-quality basic infrastructure services, such as hospitals,
schools, and water supply and sanitation, raises living standards and
increases employability of populations and prospects for inclusive growth.


Increased expenditure on infrastructure projects has a short-run effect
on employment creation as more workers are hired to build infrastruc-
ture. These jobs last only during the investment phase of the project, and,
without a continuous injection as in a stimulus-type program, such jobs
will be temporary. However, the investment program will have created a
larger stock of infrastructure capital and this permanent addition facili-
tates additional growth in the economy. The extra demand from this
incremental growth creates more jobs, and these tend to be permanent.
Furthermore, an employment experience in an infrastructure-related
employment program, even if temporary, might improve the chance of
being re-employed at a later date. This study capitalizes on the World
Bank’s long-standing knowledge on infrastructure, employment, and
growth and applies it to the case of MENA to assess the employment
creation potential of infrastructure investment.




Introduction 3


Background Information


MENA countries have been investing in infrastructure over the years.
Both in the 1990s and 2000s, public investment spending in MENA was
higher than in most developing regions, largely because of robust spend-
ing in the oil exporting countries, which benefited from rising fuel prices
(figure 1.1). Spending on infrastructure boosted employment in the


Figure 1.1 Public Gross Fixed Capital Formation
averages, percentage of GDP


0


5


10


15


20


25 a. International comparison


Ea
st


As
ia


an
d


Pa
cif


ic


Eu
ro


pe
an


d


Ce
nt


ra
l A


sia


La
tin


A
m


er
ica


an
d


Ca
rib


be
an


M
id


dl
e E


as
t a


nd


No
rth


A
fri


ca


Or
ga


ni
sa


tio
n


fo
r


Ec
on


om
ic


Co
-O


pe
ra


tio
n


an
d


De
ve


lo
pm


en
t


So
ut


h
As


ia


Su
b-


Sa
ha


ra
n


Af
ric


a


Pe
rc


en
t


b. Regional comparison


0


5


10


15


20


25


GCC Oil
importers


Developing
oil exporters


MENA


Pe
rc


en
t


1990s 2000s


Source: World Bank 2011b.
Note: GCC = Gulf Cooperation Council; MENA = Middle East and North Africa. Numbers are weighted averages of
data from the International Monetary Fund/International Financial Statistics (IMF/IFS) for a balanced sample of
countries in each region.




4 Infrastructure and Employment Creation in the Middle East and North Africa


construction sector, which was a major source of job growth in the 2000s
relative to other sectors and other countries (figure 1.2). Maintaining
momentum in infrastructure spending will be important to keep growth
and job creation from receding.


The Gulf Cooperation Council (GCC) group has the best infrastruc-
ture endowments and services in the region, reflecting high-income levels
and commitment to infrastructure investments financed by oil revenues.
Although the GCC economies have had a massive build up of infrastruc-
ture in the last decade or so, infrastructure deficiencies in developing
MENA remain a concern. Public investment spending has been particu-
larly weak in oil importing countries (OICs), which have much more
limited fiscal space than the oil exporters (figure 1.3). Although public
investment rates increased in the oil exporters in the 2000s relative to the
1990s, the opposite happened in the OICs (figure 1.1).


Recent growth in public-private partnerships was beginning to fill the
gap in some OICs. However, the economic consequences of the Arab
uprisings, combined with economic difficulties in Europe, have strained


Figure 1.2 Sectors’ Contribution to Annual Employment Growth in the 2000s


–40


–20


0


20


40


60


80


100


120


140


Typical MENA Country BrazilMalaysia


Pe
rc


en
t


Indonesia


Mining and utilities


Construction


Financial and real estate services


Agriculture


Manufacturing


Trade, tourism, logistics, and communication


Government administration and social services


Source: World Bank 2011b.
Note: MENA = Middle East and North Africa. Calculations rely on International Labour Organization data for em-
ployment by sector and Global Trade Analysis Project and United Nations Statistics Division data on value added
by sector.




Introduction 5


Figure 1.3 Fiscal Space Indicators


–20


–10


0


10


20


30


40


50


60


70


80


2008 2010 2008 2011 2008 Forecast
2011


2008 Forecast
2011


Government debt
as % of GDP


External debt as
% of GDP


Gross official
reserves, in


months of imports


Fiscal balance as %
of GDP


GCC countries Developing oil exporters Oil importers


Sources: World Bank data, IMF, and government sources.
Note: FY = fiscal year; GCC = Gulf Cooperation Council; GDP = gross domestic product.


fiscal budgets in developing MENA and reduced private investment, with
possible negative consequences for infrastructure spending. Thus, going
forward, it will be critical to protect public investment budgets and increase
resources for infrastructure projects in the case of oil importers. Doing so
will be a smart choice for governments looking to create jobs and growth.


Scope and Structure of the Study


This study consists of five chapters. Following the introductory chapter
is chapter 2, which provides an overview of the current labor market
structure in MENA as well as a summary of the future challenges the
region faces in providing employment options for its citizens. It then
gives an overview of the status of MENA’s infrastructure endowments
and calculates the domestic needs for infrastructure over the coming
decade. Based on the current share of labor used in the infrastructure
sector, the chapter finally assesses the general potential for the sector to
assist in job creation in the region.


Chapter 3 assesses the short-run impact of infrastructure investments
on job creation—direct, indirect, and induced. The chapter examines
various ways of doing this, drawing on project information as well as
adapting and using available input-output tables. For this purpose, the




6 Infrastructure and Employment Creation in the Middle East and North Africa


chapter takes advantage of a comprehensive study on infrastructure-
employment linkages in Latin America and the Caribbean and adapts its
findings to the MENA region.


Chapter 4 examines the long-term impact of infrastructure invest-
ments on economic growth and employment generation. It employs
infrastructure- and employment-growth elasticities in order to estimate
the potential benefits of infrastructure investments for employment cre-
ation between 2010 and 2020.


Chapter 5 discusses the issues faced by policy makers when consid-
ering the use of subsidized employment schemes as part of infrastruc-
ture investment programs. The chapter discusses in particular how
such short-term employment programs can be used to improve the
long-term employment prospects of workers through suitable training
and education. The chapter addresses both questions of incentives and
efficiency.


Definitions and Key Concepts


The MENA region is heterogeneous, so in addition to reporting regional
aggregates, this study reports statistics on three commonly used country
groupings—the members of the GCC countries, OECs, and OICs. Table
1.1 shows the distribution of the economies in MENA according to this
classification, along with their income categories: high-income (H),
higher middle-income (HM), lower middle-income (LM), and low-
income (L). Developing MENA refers to the OICs and OECs.


For detailed analysis of job creation, it is important to distinguish
between the jobs created within the sector where the initial investment
takes place, and those jobs created outside the sector but linked to the


Table 1.1 MENA Classification


Developing oil importers Developing oil exporters Gulf Cooperation Council


Economy
Income
bracket Economy


Income
bracket Economy


Income
bracket


Djibouti L Algeria HM Bahrain H
Egypt, Arab Rep. LM Iran, Islamic Rep. HM Kuwait H
Jordan LM Iraq LM Oman H
Lebanon HM Libya HM Qatar H
Morocco LM Syrian Arab Republic LM Saudi Arabia H
Tunisia LM Yemen, Rep. L United Arab Emirates H
West Bank and Gaza LM


Note: H = high-income; HM = higher middle-income; L = low-income; LM = lower middle-income.




Introduction 7


original investment. Conventionally, three categories are distinguished
when analyzing the employment effects of investment.


Direct employment refers to the employment created within a given
sector as it responds to an increase in the final demand for its product
such as the employment created to manufacture a wind turbine and then
to operate it. The ratio of the number of direct jobs created to the incre-
ment in spending is termed the direct employment multiplier.


Indirect employment refers to the employment created as other sectors
expand their outputs in order to supply the inputs required to produce
the output of the given sector, and the employment created by yet other
sectors as they respond to the demand for their outputs from the sectors
supplying the given sector. For example, the manufacture of steel to sup-
ply turbine components creates indirect employment, as does the extra
energy required to produce the steel. The ratio of indirect plus direct jobs
to direct jobs is named type I multiplier, and the sum of direct plus indi-
rect jobs is termed type I jobs.


Induced employment refers to the employment created to meet the
extra demand created by the additional household income of those
benefiting directly and indirectly from the initial increase in final
demand for the given sector. The extra workers in the turbine industry,
the steel industry, and the power sector spend part of their incomes on
a whole range of goods, thus creating extra employment in these sectors,
and this creates yet further spending and employment from these
incomes. The fraction of extra income spent on goods and services is a
crucial parameter, which depends on the tax and savings rates applica-
ble to the workers benefiting directly and indirectly by the initial invest-
ment. The ratio of direct plus indirect plus induced jobs to direct jobs
is called type II multiplier, and the sum of the three categories is termed
type II jobs.


For each of these categories of employment, the time dimension is
important. Investment spending generates jobs in construction, installa-
tion, and manufacture (CIM) as well as possible jobs in operations and
maintenance (O&M). The duration of the CIM jobs depends on the
nature of the investment, whereas the duration of O&M jobs is linked to
plant life. Furthermore, some jobs will be full time and some part time.
Conventionally, jobs are converted to full-time equivalents (FTEs) and
expressed in terms of job-years, so as to allow for the possibility that a
given investment expenditure will generate jobs stretching over a num-
ber of years. This aspect of employment calculation requires careful
specification of the way in which the expenditure is to be made.




8 Infrastructure and Employment Creation in the Middle East and North Africa


The calculation of long-term job creation through the links from
infrastructure to growth and growth to employment encompasses all
three types of employment, as it links total employment to total infra-
structure spending.


References


World Bank. 2004. Unlocking the Employment Potential in the Middle East and
North Africa. Washington, DC: World Bank.


———. 2007. World Development Report: Development and the Next Generation.
Washington, DC: World Bank. http://www-wds.worldbank.org/external/
default/WDSContentServer/WDSP/IB/2006/09/13/000112742_20060913
111024/Rendered/PDF/359990WDR0complete.pdf.


———. 2011a. World Development Report: Conflict, Security and Development.
Washington, DC: World Bank. http://wdr2011.worldbank.org/fulltext.


———. 2011b. Economic Developments and Prospects Report, Middle East and
North Africa: Investing for Growth and Jobs. Washington, DC: World Bank.
http://siteresources.worldbank.org/INTMENA/Resources/World_Bank_
MENA_Economic_Developments_Prospects_Sept2011.pdf.




9


C H A P T E R 2


The State of Employment and
Infrastructure and Future Needs


This chapter provides an overview of the current employment situation
in the Middle East and North Africa (MENA) region, employment in
infrastructure, and assessment of the region’s infrastructure endowments,
as well as future domestic infrastructure needs. These assessments high-
light the magnitude of the challenges faced by governments in the region
and enable us to provide a rough estimate of the scope of infrastructure
investment in solving the region’s unemployment problem in the coming
decade.


Employment Challenges


The supply and demand of workers and jobs are the underlying factors
that define labor market dynamics in an economy. Employment levels
indicate the actual demand for workers in a regional economy, whereas
the labor force numbers show the potential demand for jobs among the
population of the region. Hence, unemployment exists where the labor
force exceeds employment. Table 2.1 shows the growth trends in
employment and labor force sizes in the MENA region between 2000
and 2009.1




10 Infrastructure and Employment Creation in the Middle East and North Africa


Table 2.1 shows that, in spite of approximately 30 million jobs being
created during the past decade, the size of the labor force has concur-
rently grown by a little more than 30 million leading to an increased gap
between labor force supply and available jobs—a gap that is widening at
the rate of about 0.33 percent per year. While the number of unemployed
increased, labor force participation increased too, so the region-wide
unemployment rate declined by 3 percentage points from 13 percent in
2000 to 10 percent in 2009.


Employment creation in the MENA region needs to accelerate dur-
ing the next decade to bring down unemployment further and accom-
modate new entrants into the labor force. Less than 3 million jobs were
created annually in the 2000s, but the region should have created
1 million more jobs to bring down unemployment rates between 4 and
6 percent—a range prevalent in fast-growing economies. This is a more
modest estimate than the one presented in a 2004 World Bank report,
which argued that the labor force of the region would increase from
104 million workers in 2000 to 146 million by 2010 and then to
185 million by 2020 (World Bank 2004). The report concluded that, to
cope with unemployment and growth in the labor force, almost
100 million new jobs would have to be created over the next two
decades, amounting to 5 million jobs a year or 2 million more jobs than
the ones created in the 2000s.


The more modest estimate of 1 million additional jobs in the 2000s
suggests that the goal of solving the problem of unemployment is within
reach. Of course, achieving this regional goal would depend on develop-
ments in a few major economies. Labor force levels by country in 2009
(figure 2.1) indicate that the Arab Republic of Egypt and the Islamic
Republic of Iran, followed by Algeria and Morocco, have the largest labor
forces. Thus, labor market developments in Egypt and the Islamic


Table 2.1 Employment and the Size of the Labor Force in MENA


Region


Employment
(thousands)


Labor force
(thousands)


2000 2009 2000 2009


Gulf Cooperation Council (GCC) countries 10,995 15,473 11,511 16,387
Developing oil exporting countries (OECs) 39,588 54,742 47,187 61,598
Oil importing countries (OICs) 37,168 47,323 41,954 52,884
Total MENA 87,751 117,538 100,653 130,869


Source: International Labour Organization (ILO).
Note: MENA = Middle East and North Africa.




The State of Employment and Infrastructure and Future Needs 11


Republic of Iran have a much greater impact on regional outcomes than
other countries in the region.


Infrastructure Endowments and Future Needs


MENA has one of the highest levels of infrastructure endowments
among developing regions (table 2.2). The region, taken as a whole,
has the best access to electricity, and it has the second best paved road
network and water and sanitation systems in the developing world.
The region’s electricity generating capacity is comparable to that of
the East Asia and Pacific region and South Asia.


Although the infrastructure investment in the overall region has been
strong, intraregional differences in infrastructure endowments and quality
of infrastructure services are substantial and in line with differences in per
capita incomes. The Gulf Cooperation Council (GCC) group has the best
infrastructure endowment in the region and by far the best provision of
paved road network, telephone density, and electricity generating capac-
ity. For example, the density of the paved road network in the GCC
group is 27 times greater than that in the developing oil exporting coun-
tries (OECs), and five times more than that of the developing oil import-
ing countries (OICs) (table 2.3). When it comes to access to basic
services, the gaps among the three groups are much smaller, although it


Figure 2.1 Labor Force Levels in MENA, 2009


0 5,000 10,000 15,000 20,000 25,000 30,000 35,000


Algeria
Bahrain


Egypt, Arab Rep.
Iran, Islamic Rep.


Iraq
Jordan
Kuwait


Lebanon
Libya


Morocco
West Bank and Gaza


Oman
Qatar


Saudi Arabia
Syrian Arab Republic


Tunisia
United Arab Emirates


Yemen, Rep.


Thousands


Source: International Labour Organization (ILO).
Note: MENA = Middle East and North Africa.




12 Infrastructure and Employment Creation in the Middle East and North Africa


is worth noting that basic services have slightly higher penetration rates
in OICs compared with OECs.


Income and population growth, and baseline endowments are key
determinants of the demand for infrastructure services and investment
needs, going forward. Investment in infrastructure should, at least,
maintain infrastructure services to satisfy the demand of consumers and


Table 2.3 Infrastructure Endowments in MENA by Country Grouping
averages, 2005–08


Sector OIC OEC GCC


Density of paved road network
3,220 618 16,907 km per 1,000 km2 of arable land


Telephone density
535 538 1,351 Fixed and mobile subscribers per 1,000 people


Electricity generating capacity
0.3 0.4 2.9 Million kWh per million people


Access to electricity
98 85 98 % of population with access


Improved water
93 79 97 % of population with access


Improved sanitation
84 81 99 % of population with access


Source: World Bank Development Indicators.
Note: GCC = Gulf Cooperation Council; km2 = kilometers squared; kWh = kilowatts per hour; MENA = Middle East
and North Africa; OEC = developing oil exporting country; OIC = oil importing country.


Table 2.2 Infrastructure Endowments in the Developing World
averages, 2005–08


Sector EAP ECA LAC MENA SA SSA


Density of paved road network
1,128 1,051 2,965 2,179 467 1,095 km per 1,000 km2 of arable land


Telephone density
400 929 839 537 353 273 Fixed and mobile subscribers per 1,000 people


Electricity generating capacity
0.30 0.92 0.44 0.30 0.31 0.11 Million kWh per million people


Access to electricity
62 n.a. 86 91 48 31 % of population with access


Improved water
81 94 91 88 82 67 % of population with access


Improved sanitation
62 90 78 83 55 33 % of population with access


Source: World Bank Development Indicators.
Note: EAP = East Asia and Pacific; ECA = Europe and Central Asia; km2 = kilometers squared; kWh = kilowatts per
hour; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; n.a. = not applicable;
SA = South Asia; SSA = Sub-Saharan Africa.




The State of Employment and Infrastructure and Future Needs 13


producers, including the expansion of access to basic services, because of
population growth. This section presents estimates of the investment
needs in infrastructure for the decade starting after 2010 by considering
determinants of demand for infrastructure services for the three main
groups of countries in MENA.


The assessment of future infrastructure needs is based on econometric
models of demand for each infrastructure subsector following Fay and
Yepes (2003) and Ianchovichina et al. (2012). The models are estimated
on a worldwide dataset, although with a partial coverage of regions,
including MENA (see annex 2A for econometric results and annex 2B for
data sources). The database used for these estimations is an annual panel
data of infrastructure stocks, macroeconomic variables, and demographic
characteristics. Data for MENA countries cover years up to 2008. The
data are taken mainly from the World Bank’s World Development
Indicators (WDI) complemented with material from country official
statistical offices and other multilateral organizations.2 The annual gross
domestic product (GDP) growth rate for the world as a whole is assumed
to be 3.5 percent per year in the period 2011–20, whereas that for
MENA is 4.3 percent. OICs are assumed to grow at 3.5 percent per year,
whereas OECs and GCC economies are assumed to grow at 4.6 percent
and 4.5 percent, respectively.3 Demographic trends are taken from the
2009 United Nations World Urbanization Prospects.


This assessment improves upon earlier ones conducted for the MENA
region by (1) including data on infrastructure stocks for countries of the
region from national sources as compared to relying on extrapolations
from international databases, (2) updating the data from original interna-
tional sources, and (3) including high-income GCC economies in the
definition of MENA.


The infrastructure needs are estimated based on the levels of infra-
structure endowments required to meet household and firm demands.
The approach relies on an econometric estimate of this future demand
for infrastructure. Future demands depend on technology, the real price
of infrastructure services, per capita income, and the share of GDP
derived from agriculture and manufacturing sectors.4 Projected levels of
infrastructure stocks are valued at the unit costs used in Ianchovichina
et al. (2012) and shown in table 2.4.


To assess the full budgetary allocation needed by the various
sectors, estimates of the associated maintenance needs are included as
well. Maintenance is needed for any investment to meet its assumed
lifetime, so a commitment to maintenance is built into the cost-benefit




14 Infrastructure and Employment Creation in the Middle East and North Africa


analysis and the calculation of the social rate of return on the invest-
ments. In this study, maintenance costs are estimated by multiplying
the stock value in the previous period by a depreciation rate. A fixed
annual depreciation rate is assumed for each sector and shown in table
2.4. For paved and unpaved roads, the depreciation rate comes from a
stylized model of a maintenance and rehabilitation program.5 It
assumes roads of good quality undergo routine maintenance every year
except for year 5 and 15 when periodic maintenance and rehabilita-
tion takes place, respectively. Consequently, every year, 85 percent of
a country’s road network in good or fair condition would be subjected
to routine maintenance, 12 percent would undergo periodic mainte-
nance, and the last 3 percent would be rehabilitated. The model
assumes that roads in bad condition undergo a special rehabilitation
program to reduce their percentage to zero over a 10-year span.6
However, not all roads in bad condition will automatically be
improved; some will remain in bad condition while the program
improves the network as a whole. The annual average maintenance
and rehabilitation costs for the period between 2011 and 2020 are
then divided by the unit cost of full replacement to obtain the depre-
ciation rate as shown in table 2.5. This way, the life of the asset and
the cost of keeping its service level are incorporated into the deprecia-
tion rate. Estimates of total annual expenditure needs by sector,


Table 2.4 Unit Costs of Infrastructure by Sector
U.S. dollars


Sector Cost per unit Unit Depreciation rate (%)


Electricity generation 2,000 kW 4.0
Paved roads 410,000 km 4.7
Unpaved roads 50,000 km 7.2
Rail lines 900,000 km 4.0
Rural water and sanitation 150 person 3.0
Urban sanitation 150 person 3.0
Rural sanitation 130 person 3.0
Urban water 80 person 3.0
Main telephone linesa 127–580 line 8.0
Mobile linesa 127–451 line 8.0
Access to electricity 195 person 4.0
Ports 348 TEU 4.0
Wastewater treatment 120 person 3.0


Source: Ianchovichina et al. 2012.
Note: km = kilometers; kW = kilowatt.
a. Varies by region.




The State of Employment and Infrastructure and Future Needs 15


subgroup, and region are presented in table 2.6. In the absence of data
for a sector or subsector in a country, regional averages of investment
were imputed to obtain total investment needs (see annex 2C).


Annual additional infrastructure investment and maintenance needs
through 2020 for the MENA region are estimated to be US$106 billion
or 6.9 percent of the regional GDP. OECs have the highest demand,
representing 46 percent of the regional needs, and they will need to com-
mit almost 11 percent of their GDP annually (US$48 billion) to keep up
with their economic and demographic growth. The needs of OICs and
the GCC economies are smaller at approximately 6 and 5 percent,
respectively, of their GDP.


Investment and rehabilitation needs are particularly high in the trans-
port and electricity sectors. These account for 44 and 45 percent of total
needs, respectively, followed by information and communication tech-
nology (ICT), and water and sanitation. During the next 10 years, OECs
will need to spend 5.1 percent of their GDP (or US$22.5 billion) annu-
ally on upgrading and expanding their transport networks, whereas OICs
and GCC will need to spend around 2.0 percent of their respective GDP.


Approximately 40 percent of infrastructure expenditure meets capital
expansion needs, whereas the remainder is allocated to infrastructure
maintenance projects (figure 2.2). Of the US$46 billion that should be
spent annually in the transport sector to keep up with total demand
(table 2.6), about US$30 billion must be earmarked for maintenance as
suggested by the expenditure shares required for new capital and for
maintenance, as reported in table 2.7.


The vast share of infrastructure expenditure is expected to fund projects
that meet the needs for transport and electricity infrastructure. Within


Table 2.5 Roads Maintenance and Rehabilitation Program
U.S. dollars


Component


Capital intensive technology


Paved roads Unpaved roads


Unit costs
Routine 4,000 2,000
Periodic 50,000 10,000
Rehabilitation 150,000 15,000
Maintenance and rehabilitation (annual cost per km) 19,100 3,600
Unit costs for new roads 410,000 50,000
Depreciation rate (%) 4.7 7.2


Source: World Bank data.
Note: km = kilometers.




16 Infrastructure and Employment Creation in the Middle East and North Africa


transport, paved roads are expected to absorb most of the funds and
account for more than half of the total transport infrastructure needs in
OECs (50 percent), almost two-thirds of the needs in OICs (63 percent),
and less than one-third of the needs in GCC countries (28 percent).


Table 2.6 Annual Expenditure Needs for Infrastructure in the MENA Region


Sector OIC OEC GCC Total


Percentage of GDP
Transport 2.2 5.1 2.0 3.0
Paved roads 1.4 2.7 0.6 1.4
Unpaved roads 0.6 2.1 1.3 1.4
Rail lines 0.1 0.1 0.0 0.1
Ports 0.1 0.1 0.1 0.1
ICT 0.8 1.1 0.2 0.6
Telephone mainlines 0.2 0.3 0.0 0.2
Mobile lines 0.6 0.7 0.2 0.4
Electricity 2.5 4.2 2.5 3.0
Electricity generation 2.1 3.7 2.4 2.7
Electricity access 0.4 0.5 0.1 0.3
Water and sanitation 0.5 0.6 0.1 0.3
Water 0.2 0.2 0.0 0.13
Sanitation 0.3 0.3 0.1 0.19
Total 6.0 10.9 4.8 6.9


Amount, 2005 US$, million
Transport 8,575 22,492 14,453 45,519
Paved roads 5,448 12,088 3,956 21,492
Unpaved roads 2,246 9,497 9,461 21,204
Rail lines 428 639 75 1,143
Ports 452 268 960 1,680
ICT 3,021 4,707 1,559 9,287
Telephone mainlines 696 1,464 259 2,419
Mobile lines 2,325 3,243 1,300 6,868
Electricity 9,894 18,607 17,602 46,103
Electricity generation 8,214 16,467 17,139 41,820
Electricity access 1,680 2,140 463 4,283
Water and sanitation 1,764 2,497 647 4,908
Water 745 1,040 190 1,975
Sanitation 1,019 1,458 457 2,934
Total 23,254 48,303 34,261 105,818
Share by country group 22% 46% 32% 100%
Investment 10,261 20,739 15,786 46,786
Maintenance 12,992 27,564 18,475 59,032


Source: World Bank data.
Note: GCC = Gulf Cooperation Council; GDP = gross domestic product; ICT = information and communication
technology; MENA = Middle East and North Africa; OEC = developing oil exporting country; OIC = oil importing
country.




The State of Employment and Infrastructure and Future Needs 17


Electricity is the sector with the largest total expenditure needs.
Annual needs for electricity amount to US$46 billion per year in MENA,
or 3 percent of the regional annual GDP (table 2.6). Ensuring adequate
electrification access for 2011–20 is estimated to require US$4 billion per
year, whereas the bulk of the annual expenditure (US$42 billion) will be
needed to secure adequate levels of generation capacity.


Expenditure needs for ICT will amount to around $9 billion per
year between 2011 and 2020 and reflect mostly expenditure on mobile
communication infrastructure (table 2.6). The bulk of this expenditure


Figure 2.2 Composition of Infrastructure Expenditure Needs by Group
of Countries


a. Type of expenditure


b. Sectors


100


80


60


40


20


0


100


80


60


40


20


0


OICs OECs


OICs OECs GCC MENA


GCC


Pe
rc


en
t


Pe
rc


en
t


MENA


Capital expansion Capital replacement


Transport ICT Electricity Water and sanitation


Source: World Bank data.
Note: GCC = Gulf Cooperation Council; ICT = information and communication technology; MENA = Middle East
and North Africa; OEC = developing oil exporting country; OIC = oil importing country.




18 Infrastructure and Employment Creation in the Middle East and North Africa


will fund ICT investments in developing MENA, and nearly two-thirds of
this expenditure will be spent on maintenance projects (table 2.7).


Annual water and sanitation expenditure needs amount to US$4.9
billion per year, or 0.3 percent of regional GDP, with US$2.9 billion needed
to meet demand for water and sanitation services in urban areas (see
table 2.8). The Millennium Development Goals (MDGs) were taken into
account when estimating the investment needs for water and sanitation.


Table 2.8 Expenditure Needs for Water and Sanitation in Urban and Rural Areas


Subsector


OIC OEC GCC Total


% of
GDP


US$,
millions


% of
GDP


US$,
millions


% of
GDP


US$,
millions


% of
GDP


US$,
millions


Urban areas 0.23 888 0.36 1,613 0.06 403 0.19 2,904
Water 0.08 308 0.13 577 0.02 141 0.07 1,025
Sanitation 0.15 580 0.23 1,037 0.04 262 0.12 1,879
Rural areas 0.22 876 0.20 884 0.03 244 0.13 2,004
Water 0.11 437 0.10 463 0.01 49 0.06 949
Sanitation 0.11 439 0.09 421 0.03 195 0.07 1,055


Source: World Bank data.
Note: GCC = Gulf Cooperation Council; GDP = gross domestic product; OIC = Oil importing country;
OEC = Developing oil exporting country.


Table 2.7 Annual Infrastructure Investment and Maintenance Needs in MENA
by Type of Investment as Percent of GDP


Sector


OIC OEC GCC Total


N M N M N M N M


Transport 0.8 1.4 1.8 3.3 0.7 1.4 1.0 1.9
Paved roads 0.6 0.8 1.1 1.6 0.3 0.3 0.6 0.8
Unpaved roads 0.2 0.4 0.6 1.5 0.4 1.0 0.4 1.0
Rail lines 0.0 0.1 0.0 0.1 0.0 0.0 0.0 0.1
Ports 0.1 0.0 0.0 0.0 0.1 0.1 0.1 0.0
ICT 0.26 0.51 0.42 0.65 0.06 0.16 0.21 0.39
Telephone mainlines 0.08 0.10 0.15 0.18 0.02 0.02 0.07 0.09
Mobile lines 0.18 0.41 0.27 0.47 0.04 0.14 0.14 0.30
Electricity 1.37 1.16 2.22 1.98 1.44 1.04 1.65 1.34
Electricity generation 1.22 0.88 2.04 1.67 1.42 1.00 1.55 1.16
Electricity access 0.15 0.28 0.17 0.31 0.02 0.04 0.10 0.18
Water and sanitation 0.19 0.27 0.28 0.29 0.04 0.05 0.14 0.17
Water 0.07 0.12 0.12 0.12 0.01 0.02 0.06 0.07
Sanitation 0.11 0.15 0.16 0.17 0.03 0.04 0.09 0.10
Total 2.63 3.33 4.68 6.22 2.23 2.61 3.03 3.83


Source: World Bank data
Note: GCC = Gulf Cooperation Council; GDP = gross domestic product; ICT = information and communication
technology; M = maintenance; MENA = Middle East and North Africa; N = new capital; OIC = Oil importing
country; OEC = Developing oil exporting country.




The State of Employment and Infrastructure and Future Needs 19


Analyzing trends in access to these services show that most countries are
behind schedule in achieving these by 2015—only Egypt and Qatar are on
track. Table 2.9 shows the access shortfall by country and country group,
defined as additional access, which should have been completed by 2008,
if countries were following a linear achievement path toward the MDGs.


Potential of Infrastructure Investment
in Boosting Employment


This section provides a first-round assessment of the extent to which infra-
structure activities could play a role in speeding up job creation and
resolving MENA’s employment problem. It presents employment shares
by infrastructure sector and country and compares those to international
benchmarks. The data used in this section summarize the information that
can be computed based on the sectoral disaggregation of employment fol-
lowing the international standard industrial classification (ISIC) of sectors,
allowing an identification of electricity and water sector jobs as well as
jobs in transport and communication sectors.7 These activities add up to


Table 2.9 Access Shortfall Compared to MDG Linear Path Achievement
percentage points


Country/economy


Urban areas Rural areas


Water Sanitation Water Sanitation


Oil importing country
Djibouti 19.7 28.2 54.8
Jordan 1.4 0.7 3.2
Morocco 4.8 11.2 1.3
Tunisia 0.8 0.2
West Bank and Gaza 9 2.6 4.6
Oil exporting country
Algeria 15 1.4 13.3
Iran, Islamic Rep. 0.7 4.1 3.3 5.8
Iraq 7.1 6.9 9.2
Libya 1.8 1.1 2.6 1.4
Syrian Arab Republic 3.4 0.2 0
Yemen, Rep. 19.5 14.6 6.8
GCC
Kuwait 0.4 0.4
Oman 1.1 5.1 0.7
Saudi Arabia 1.1 0.1
United Arab Emirates 0.7 1.8


Source: World Bank data.
Note: GCC = Gulf Cooperation Council; MDG = Millennium Development Goal. There are no data available for
rural sanitation for Saudi Arabia. A blank space means the country is on track to achieve the relevant MDG.




20 Infrastructure and Employment Creation in the Middle East and North Africa


what is usually defined as infrastructure in the sector. In spite of the lim-
ited ability to unbundle employment into subsector specific details, these
data prove extremely useful as they allow identification of jobs associated
with the delivery of electricity and water services as well as a really good
approximation of jobs in transport and communication services.


Data on construction are included in this section as an important ele-
ment of infrastructure investment.8 The construction category encom-
passes housing and building construction. These two activities are likely to
be the main drivers of job creation, but other infrastructure investments
are still likely to account for a significant share of employment creation.


The infrastructure and construction sectors in the region employ
approximately 18.2 million people—10.6 million in construction and
7.6 million in infrastructure. About 2.2 million of these jobs are in the
GCC economies, 9.2 million in other MENA OECs, and 6.8 million in
the MENA OICs. On average, jobs in the infrastructure sector represent
around 8.0 percent of the employment in the region, whereas construc-
tion jobs constitute around 11.3 percent (table 2.10).


OECs have a higher share of jobs in the infrastructure and construc-
tion sectors, not only relative to other groups in the region but also to
international benchmarks. In contrast, oil importers’ employment shares
are more in line with international averages, although their employment
in infrastructure is below various international benchmarks (table 2.10).
It is worth noting that many of the infrastructure and construction jobs
in the GCC economies are performed by migrant workers, although lack
of data precludes separating the impact of infrastructure investments on
employment of migrant and national workers.


Table 2.10 Employment Shares of Infrastructure and Construction Sectors
Percentage of all employed in 2009


Infrastructure
services share of total


employment


Construction
share of total
employment


Total infrastructure
services and


construction share


GCC 5.8 12.0 17.7
OECs 9.6 13.0 22.6
OICs 7.1 9.4 16.5
Total for MENA 8.0 11.3 19.3
World average 7.7 8.4 16.1
Developed countries 7.2 8.2 15.4
Developing countries 7.6 8.6 16.2


Sources: World Bank and ILO data.
Note: GCC = Gulf Cooperation Council; MENA = Middle East and North Africa; OEC = developing oil exporting
country; OIC = oil importing country.




The State of Employment and Infrastructure and Future Needs 21


Assuming that world benchmarks gauge the normal size of employ-
ment in the infrastructure sector (table 2.10), we conclude that in general
there is limited scope for infrastructure activities to speed up job creation.
In the OECs, infrastructure’s share in total employment is much higher
than the world average, so the scope for increasing the relative size of
employment in the sector is limited. The same is the case with employ-
ment in construction in all subregions and the region as a whole, although
there is some scope to expand the relative size of employment in infra-
structure services in oil importing and GCC countries, where infrastruc-
ture’s employment share is below world norms.


Within infrastructure, the transport and communication sectors are
the most labor intensive, employing about 7 percent of all employed,
compared to approximately 1 percent in electricity and water sectors
(table 2.11). This means that in 2009, the electricity and water sectors in
the region relied on a labor force of about 1 million workers, whereas
transport and communications employed approximately 6.5 million
workers. GCC’s employment levels within infrastructure are significantly
lower than international averages. In contrast, developing oil exporters
have a significantly larger employment share in the transport and com-
munication subsector than that of other countries. Their employment
shares in the electricity and water sectors are consistent with those of
other regions, and there is some scope for an increase when compared
with the average standards of developing countries.


These aggregate numbers hide significant variations across economies.
For example, figure 2.3 shows that the Islamic Republic of Iran is an
outlier in terms of employment in the construction sector as it employs
almost four times as many workers as the regional average. Egypt and
Saudi Arabia stand out as the only two economies employing more work-
ers in the infrastructure sector than in the construction sector.


Table 2.11 Infrastructure Jobs by Sector in MENA
percent


Electricity and water Transport and communication


GCC economies (GCC) 0.8 4.9
Developing oil exporting


countries (OECs)
1.1 8.5


Oil importing countries (OICs) 1.1 6.0
Total for MENA 1.1 6.9
Developed countries 0.9 6.3
Developing countries 1.2 6.6


Source: Based on ILO data.
Note: GCC = Gulf Cooperation Council; MENA = Middle East and North Africa.




22 Infrastructure and Employment Creation in the Middle East and North Africa


This overview of the relative importance of the infrastructure-related
sectors for employment in the region provides a reality check. It suggests
that investment in infrastructure alone will not solve the employment
problem in the region. According to the employment estimates discussed
in this chapter, the infrastructure sector would have to increase its
employment by more than 20 percent every year to meet the overall
annual job creation target of the region. Such a high rate of growth is
difficult to sustain over the long run, given the already large relative size
of employment in infrastructure and construction in the region.


However, infrastructure investments can certainly play a significant role
in efforts to create jobs in the region. Annual infrastructure investment
and maintenance needs through 2020 are estimated to be $106 billion or
6.9 percent of the regional GDP. OECs have the highest demand, repre-
senting 46 percent of the regional needs, and they will need to commit
almost 11 percent of their GDP annually to keep up with their economic
and demographic growth. The needs of the OICs and the GCC economies
are smaller at 6 and 5 percent of GDP, respectively. Next, chapters 3 and
4 discuss the implications of infrastructure investment for employment
creation in the short and medium run, respectively.


Figure 2.3 Shares of Infrastructure and Construction Jobs in Total Employment
in MENA


Total Construction Infrastructure


0 10 20 30 40 50
Percentage of total jobs in the country


Algeria


Bahrain


Egypt, Arab Rep.


Iran, Islamic Rep.


Iraq


Jordan


Kuwait


Morocco


West Bank and Gaza


Oman


Qatar


Saudi Arabia


Syrian Arab Republic


Tunisia


United Arab Emirates


Yemen, Rep.


Source: Based on ILO data.
Note: MENA = Middle East and North Africa.




23


Annex 2A Econometric Models for Infrastructure Needs


Paved roads Total roads Rail lines Ports Telephone mainlines Mobile lines
Electricity
generation


Method Probit for grouped data Fixed effects Fixed effects Fixed effects Logit for grouped data Logit for grouped data Fixed effects


Per capita GDP −0.261***
(0.00189)


0.111*
(0.0575)


−0.0107
(0.0422)


1.124***
(0.1640)


0.983***
(0.0002)


0.414***
(0.0002)


0.652***
(0.0549)


Share of manufactures
in GDP


0.131***
(0.0018)


0.00209
(0.0435)


−0.0428
(0.0395)


0.152
(0.1110)


0.0686***
(0.0002)


−0.142***
(0.0002)


0.199***
(0.0343)


Share of agriculture in
GDP


−0.179***
(0.0020)


−0.0772
(0.0491)


0.0630
(0.0393)


−0.202*
(0.1210)


0.199***
(0.0002)


0.184***
(0.0002)


0.147***
(0.0429)


Population density −0.574***
(0.0046)


−0.427***
(0.1160)


−0.929***
(0.1020)


0.409
(0.2980)


−0.00914***
(0.0003)


2.133***
(0.00034)


−0.00847
(0.1000)


Urbanization −0.493***
(0.0054)


0.377***
(0.1370)


−0.0421
(0.1120)


0.279
(0.3370)


3.361***
(0.0005)


1.259***
(0.0005)


0.141
(0.0872)


Population growth n.a. n.a. n.a. n.a. −1.645***
(0.0013)


−0.828***
(0.0016)


n.a.


Time trend 0.0985***
(0.0005)


−5.684
(4.1460)


−1.224
(1.3040)


n.a. n.a. n.a. 0.0697***
(0.0126)


Time trend squared n.a. 0.00714
(0.0052)


0.00155
(0.0016)


n.a. n.a. n.a. n.a.


Market age n.a. n.a. n.a. 0.102***
(0.0121)


0.107***
(0.0092)


0.111***
(0.0114)


n.a.


Market age squared n.a. n.a. n.a. −0.00117***
(0.0002)


−0.00366***
(0.0034)


0.00414***
(0.0052)


n.a.


Constant −35.32***
(0.1910)


1,128
(828.6000)


236.2
(260.0000)


−15.28***
(2.0940)


−8.373***
(0.0028)


−13.74***
(0.00289)


−34.94***
(4.7220)


Observations a 633 551 496 a a 1,034
R2 0.119 0.540 0.801 0.4590


Number of coefficients 172 109 102 173


Source: Ianchovichina et al. 2012.
Note: GDP = gross domestic product. Standard errors in parentheses. All relevant variables are in logarithms.
a. indicates a large number of observations due to the grouped technique.


*p < .1 **p < .05 ***p < .01 n.a. = not applicable.




24 Infrastructure and Employment Creation in the Middle East and North Africa


Annex 2B Data Sources and Descriptions Used
for Model of Investment Requirements


Gross domestic product (GDP) in constant 2000 US$ is taken from the
World Development Indicators (WDI) database of the World Bank
(http://data.worldbank.org/) and the United Nations (UN) National
Accounts Main Aggregates Database (http://unstats.un.org/unsd
/snaama/selbasicFast.asp). GDP projections for all MENA economies
except for West Bank and Gaza come from the World Bank Growth
Forecasting Tool described in Ianchovichina and Kacker (2005).


Shares of value added in agriculture and manufacturing come from the WDI
database of the World Bank (http://data.worldbank.org/) and UN
National Accounts Main Aggregates Database (http://unstats.un.org
/unsd/snaama/selbasicFast.asp).


Total and urban population data are taken from the WDI database of the
World Bank (http://data.worldbank.org/). Projections are obtained
from the UN World Urbanization Prospects (2009 revision) (http://
esa.un.org/unpd/wup/index.htm).


Containerization, measured as the total number of trafficking containers,
is obtained from the WDI database of the World Bank (http://data.
worldbank.org/) and harmonized with the Containerization
International Yearbook (1970–2006), published by Containerization
International (http://www.ci-online.co.uk).


Telephone lines, mobile phones (in subscribers per 1,000 inhabitants),
paved and total roads, and rail lines (in thousands of kilometers)
come from the WDI database of the World Bank (http://data
.worldbank.org/).


Electricity generating capacity, in millions of kilowatts, is taken from the
U.S. Energy Information Administration, (http://tonto.eia.doe.gov
/cfapps/ipdbproject/IEDIndex3.cfm?tid=2&pid=2&aid=12).


Electrification rate, measured as the fraction of population with access to
electricity, is obtained from World Energy Outlook 2006, 2009, and
2010 published by the International Energy Agency (http://www
.worldenergyoutlook.org/).


Access to improved water and sanitation in urban and rural areas is defined
as the fraction of total population with access to these services. It is
taken from the WDI database of the World Bank (http://data
.worldbank.org/).




The State of Employment and Infrastructure and Future Needs 25


Waste water treatment, measured as the fraction of the population connected
to public waste water treatment plants, is obtained from the UN Statistical
Division (http://unstats.un.org/unsd/environment/wastewater.htm).


Kilometers of total and paved roads and port traffic in tons for Oman are
obtained from the Ministry of National Economy (http://www
.moneoman.gov.om/Stat_Online_desp.aspx).


Electrification rate for Djibouti is taken from the energy survey that was
carried out by the government and the World Bank (http://www
.ministere-finances.dj/statistiques/Projets/rapportfinalenergie.pdf).


Annex 2C Data Imputations


In the absence of data for a sector or subsector in a country, regional aver-
ages of investment as percent of GDP were imputed to obtain total
investment needs. Table 2C.1 shows the economies and sectors for which
regional averages were imputed.


Table 2C.1 Imputation of Average Investment as Percent of GDP When Data
Were Not Available


Sector Economies


Transport
Paved roads Syrian Arab Republic
Unpaved roads Syrian Arab Republic
Ports Bahrain, Iraq, Libya, Qatar, Syrian Arab Republic, and


West Bank and Gaza
Electricity
Electricity (access) West Bank and Gaza
Water and sanitation
Water, rural areas Bahrain
Sanitation, rural areas Bahrain and Saudi Arabia


Note: GDP = gross domestic product.


Notes


1. All the labor data are extracted from the International Labour Organization
(ILO) database, published on their website. Upon request, additional data were
provided by ILO staff. Data after 2008 are projections and not measurements.


2. In some cases, recent trends of annual growth of the same series are used to
fill in missing values. United Nations National Accounts Main Aggregates
Database is used for historical GDP and for the value-added components of




26 Infrastructure and Employment Creation in the Middle East and North Africa


agriculture and manufacturing. Additional data on container port traffic is
taken from the Containerization International Yearbook (1970–2006). Extra
information on roads and paved roads is available for Oman. GDP projections
are from the World Bank.


3. When past growth rates were not available, a Hodrick-Prescott filter (with a
smoothing factor of 100) was used to obtain the long-run growth rate trend
of the past 10 years (1999–2009), resulting in an implied annual growth rate
equivalent to that of the World Bank projections. Growth rates are then used
to obtain 2010–20 GDP scenarios.


4. Per capita stocks of infrastructure are regressed against per capita income, the
share of GDP derived from agriculture and manufacturers, demographic den-
sity, and urbanization rate. Time dummies and country fixed effects are used
to proxy differences in infrastructure prices. In the case of telecommunications
and ports, a market-age variable accounts for the speed of technological change
across countries, that is, mobile penetration and adoption of containerization.
Lagged dependant variables are included to eliminate the structural part of
interest. Analysis of spurious regressions have been made in these studies
showing that a structure of lagged variables as in Arellano and Bond estima-
tions can eliminate all variance thus eliminating the structural part of interest
(Yepes, Pierce, and Foster 2008).


5. This model is a simplified version of the Highway Design and Maintenance
Standards Model developed by the World Bank.


6. As there is no available information on road quality in the MENA countries,
global average quality levels are used as in Ianchovichina et al. (2012).
Although reaching 100 percent of good quality is an unrealistic goal, the
estimation of investment needs is based on demand. A road is expected to
deliver the service for which it was built, despite the fact that it could actually
perform at lower quality. Operation at lower quality represents either the
accumulation of a larger liability in terms of rehabilitation or costs internal-
ized by users.


7. The data are from the ILO tables 4b and 4c, which disaggregate employment
into industrial sectors at the 1-digit level, according to the more recent ISIC
of economic activities; Revision 3 (1990) in 4b and Revision 2 (1968) in 4c.
The ILO sector E covers two sectors: electricity, gas, steam, and hot water
supply (sector 40, ISIC); and collection, purification, and distribution of water
(sector 41, ISIC). ILO’s sector I covers five ISIC categories: land transport;
transport via pipelines (sector 60, ISIC); water transport (sector 61, ISIC); air
transport (sector 62, ISIC); supporting and auxiliary transport activities and
activities of travel agencies (sector 63, ISIC); and post and telecommunica-
tions (sector 64, ISIC).


8. ILO’s sector E matches the ISIC sector covering construction (sector 45,
ISIC).




The State of Employment and Infrastructure and Future Needs 27


References


Fay, M., and T. Yepes. 2003. “Investing in Infrastructure: What Is Needed from
2000 to 2010.” World Bank Policy Research Working Paper 3102, World Bank,
Washington DC.


Ianchovichina, E., and P. Kracker. 2005. “Growth Trends in the Developing World:
Country Forecasts and Determinants.” World Bank Policy ResearchWorking
Paper 3775. World Bank, Washington DC.


Ianchovichina, E., A. Estache, R. Foucart, G. Garsous, and T. Yepes. 2012. “Job
Creation through Infrastructure Investment in the Middle East and North
Africa.” Policy Research Working Paper No. 6164, World Bank, Washington,
DC.


World Bank. 2004. Unlocking the Employment Potential in the Middle East and
North Africa. Washington, DC: World Bank.


Yepes, T., J. Pierce, and V. Foster. 2008. “Making Sense of Sub-Saharan Africa’s
Infrastructure Endowment: A Benchmarking Approach.” Africa Infrastructure
Country Diagnostic Working Paper No. 1, World Bank, Washington, DC.






29


C H A P T E R 3


Short-Run Employment Effects
of Infrastructure Investment


The cost of creating a job is a crucial factor when assessing the potential
of infrastructure investments in creating employment. This cost varies
across sectors, making expenditure in certain areas more job intensive
than in others. The cost of job creation also varies across countries; a given
expenditure can create many more jobs in one country than in another.


This chapter describes different methods for assessing the employ-
ment generating potential of infrastructure investments in the short run.
The next three sections present the application of different approaches
toward assessing the short-term employment effects of infrastructure
investments in the Middle East and North Africa (MENA) region. The
last section discusses the implications of using labor-intensive technolo-
gies in the maintenance of unpaved roads.


Techniques for Estimating the Cost of a Job and the
Employment Generated by Investment in Infrastructure


Employment generated by infrastructure spending is typically assessed
in two steps. First, direct employment is estimated using information on
the nature of infrastructure spending and the amount of different types
of labor required by financed projects. Second, indirect and induced




30 Infrastructure and Employment Creation in the Middle East and North Africa


employment effects are estimated using multipliers from past experience
that link these types of employment to direct employment.


To estimate direct employment a given investment project would
create, it is necessary to start with actual data from a similar infrastructure
project. The ratio of jobs generated to investment expenditure can then
be used directly or with some adjustments to produce estimates in cases
where actual data are not available. The accounting identity connecting
output to its component parts represents the link between investment
spending and the employment generated by this investment (Scottish
Government 2011):


Output at basic prices = Total domestic purchases at basic prices +
Imports + Taxes on products + Compensation
of employees + Gross operating surplus.


The output at basic prices corresponds to the amount spent on a
project. Total domestic purchases at basic prices correspond to the pur-
chase of inputs from all sectors of the economy. Imports are items
financed by the investment and purchased directly from abroad. Taxes on
products include items, such as property taxes, capital, and payroll taxes.
The compensation of employees is the total payment to labor for produc-
ing the output, including wages and other nonwage benefits, such as
insurance and other benefits. The gross operating surplus is the remainder
accruing to companies for their production.


The total compensation of employees covers various full-time and
part-time jobs paying different wage rates. The simplest approach to link
total expenditure of a project to the direct employment it creates is to
divide total employee compensation by the average annual wage plus
benefits for the sector, to produce an estimate of the number of full-time
equivalent (FTE) employees.1 This estimate provides a simple link
between expenditure and direct jobs, that is, the employment multiplier.


A more detailed approach would be (1) to categorize jobs by skill
levels, for example, skilled, semiskilled, and unskilled; (2) to assign to each
level the average wage plus benefits for that skill level; and (3) finally, to
identify the number of FTE employees in each group, so that the sum of
the number of FTE employees in each skill group times their average
wage is equal to total employee compensation (Schwartz, Andres, and
Dragoiu 2009). This would provide three ratios between expenditure and
the employment types, or three employment multipliers for the respective
skill levels.




Short-Run Employment Effects of Infrastructure Investment 31


The employment multipliers for an actual project can be used to
obtain an estimate of direct employment generated by expenditure asso-
ciated with a similar project. There are two main sources of information
that provide evidence on employment multipliers. Project-level data that
may have been collected could give evidence on expenditures and
employment creation (REPP 2001). Project-level data have the advan-
tage of referring to a specific type of investment and, in cases where the
projection is required for exactly the same type of investment, these data
will provide the most reliable projection of likely employment. The accu-
racy of the projection declines with increase in the discrepancy between
the proposed and the specific investment project.


In many cases, project-level data are not available or are available only
for a specific technology or project that differs substantially from the one
that needs to be assessed. In these cases, the usual practice is to use data
from an input-output (IO) table (CH2MHILL 2009). The IO table, if
available, provides for each sector of the economy the average expendi-
ture on each input required to produce a unit of output of that sector.
As this includes total employee compensation, it can be used to connect
total expenditure and expenditure on employment in that sector. The IO
table needs to be supplemented by national data on sector wages or
employment in order to convert the employee compensation figure to an
employment figure per unit of expenditure. This ratio can then be
applied to future potential expenditures to generate an estimate of associ-
ated employment.2


The direct employment effect understates, in some instances signifi-
cantly, the total employment generated by infrastructure spending. In
some instances, the sum of indirect plus induced jobs created by infra-
structure spending has been estimated to be of comparable magnitude to
the number of direct jobs created by this investment. In the United States,
six different types of highway projects had type II multipliers3 averaging
1.9, whereas a transmission line power project had a type II multiplier of
1.7 (Babcock et al. 2010; Pfeifenberger et al. 2010). A study on Malaysia
by Bekhet (2011) found type II multipliers to be 3.5 for investments in
the electricity and gas sectors, 2.4 for water works and supply, and 1.7 for
building and construction. A study on the Egyptian 2008/09 stimulus
package (ILO 2010) reported infrastructure type II multipliers ranging
between 1.2 for investments in water and sewage and 1.6 for investments
in construction and building. These and other results in the literature
indicate that in addition to direct effects, it is desirable to estimate indi-
rect and induced employment effects of an infrastructure project.




32 Infrastructure and Employment Creation in the Middle East and North Africa


Type I and type II multipliers can be derived from IO tables, using the
Leontief inverse that enables the calculation of the costs associated with
the production of extra output in a given sector by factoring in all the
inputs required (directly and indirectly) for its production. As with the
direct employment multiplier, it is necessary to convert employee com-
pensation in each sector into employment, using sector-level wage plus
benefit rates. Once total employment generated by an initial investment
is estimated, it is possible to calculate the cost of creating a single job.
Box 3.1 presents a discussion of the strengths and weaknesses of using IO
tables for generating employment estimates. Many countries do not have
IO tables, but hybrid methods, such as those pioneered by Schwartz,
Andres, and Dragoiu (2009) and by LECG (2009), can be adapted to
address transferability across countries at different times and stages of
development.


Hybrid Approaches to Estimating the Short-Term
Employment Effects of Infrastructure Investment


Hybrid approaches work best in those cases where economies have
similar structures and are at similar levels of development. Hybrid
approaches for estimating employment generated by infrastructure
investment can have two components—limited hybrid case and full
hybrid case. In the limited hybrid case, direct employment generated by
investment in a country may be available from project-level data for the
same country. Indirect and induced employment are derived using mul-
tipliers from another country where there is an IO table (see Schwartz,
Andres, and Dragoiu 2009). In the full hybrid case, there would be no
relevant data for a certain country, and all information, including the
link from expenditure to direct, indirect, and induced employment, has
to be taken from data available for another country.


An important factor that must be recognized while making compari-
sons over time or across countries is that labor costs vary in nominal
terms. High labor costs will translate into high overall costs of creating a
job or equivalently one needs to recognize that a given nominal expendi-
ture in a sector generates less employment (direct and indirect) at higher
wages. Assuming a constant wage rate across a group of countries, as in
Schwartz, Andres, and Dragoiu (2009), might be acceptable when one
needs to obtain an aggregate number for a group of countries. However,
a calculation of country-level employment figures based on regional




Short-Run Employment Effects of Infrastructure Investment 33


Box 3.1


Pros and Cons of Input-Output Table Use for Generating
Employment Estimates


Strengths
The major advantage of using input-output (IO) tables for generating employ-


ment estimates is that they make it possible to calculate not only direct but also


indirect and induced employment. Project-based data can be used to generate


only direct employment estimates.


The IO approach is well suited for estimating employment effects of large sec-


tor programs that reflect the existing mix of projects because sectors in IO tables


tend to be broad aggregates, including energy, water, and sewage, or roads.


Although IO tables are often out-of-date by several years, studies have shown


that changes in IO tables over a decade make only small differences to employ-


ment multipliers.


Weaknesses
For projects that are not typical of the sector mix, the IO coefficient could result


in misleading estimates of employment generated by a given amount of


investment.


The labor intensity for a particular project could be higher or lower than the


sector average, and the use of indirect employment could be quite different from


the sector average. For example, the scale of a project could affect labor intensity so


that the sector average would not represent a project at the particular scale required.


The use of the employment multiplier implies that the wage associated with


a potential project is equal to the average wage rate for that sector. This assump-


tion may not be accurate in cases where there is considerable wage variation and


may bias the estimates of the employed and the induced employment resulting


from consumption financed by employee compensation.


IO models do not incorporate potentially important supply constraints.


Employment may not be able to respond to demand. Large-scale investments


run the risk of creating a “crowding out” effect with employment increasing in the


target sector but decreasing elsewhere. These concerns are most relevant for


large projects in countries with tight labor markets.




34 Infrastructure and Employment Creation in the Middle East and North Africa


average wages could differ substantially from the one based on country-
specific wages.4


Given these issues, adjustments have to be made to values taken
from IO tables that reflect wage rate changes over time and wage dif-
ferences across countries. Adjustments are required when there are
differences between the year of the IO table calculation and the year
of the investment-generated employment estimate. Adjustments are
also required when values have to be imported from one country with
an IO table to another without an IO table.


The general level of wages can be indexed by gross domestic product
(GDP) per capita in nominal terms to correct for increases in per capita
income and income differences across countries, as suggested by LECG
(2009). Using data from Gardiner and Theobald (2010)5 and regressing
the hourly construction wage, inclusive of benefits, on GDP per capita,
both expressed in 2009 US$ (table 3.1), one can show that wages are
highly and positively correlated to GDP per capita. This result provides
evidence that the approach of scaling down one country’s wages by the
ratio of its GDP per capita to another country’s GDP per capita is a good
approximation to wages in the other country. The adjustment, however,
would not be sufficient in cases where individual sector wages change
relative to the general wage rate. It is worth noting that correcting
employment generated in every sector by the same factor will reduce the
estimated number of jobs created (direct, type I, and type II), but will not
alter the values of the type I and type II multipliers.


A third correction in the process of deriving the direct employment
calculation involves the conversion of national data on basic wage rates to
gross rates, including benefits. Gardiner and Theobald (2010) provide
data for the basic wage in the construction sector and a total labor cost
basis for a number of countries, including the Arab Republic of Egypt.
Table 3.2 indicates that for Egypt, the ratio of basic to total wages
increases slightly with the skill level and varies between 0.83 and 0.88.
The ratio for semiskilled workers is taken as an estimate of the differential
between basic and total wages for all infrastructure employees.


Table 3.1 Regression of Semiskilled Hourly Construction Wage on GDP per Capita


Coefficient t-statistic


Intercept 1.15 0.54
GDP per capita 0.00084 12.0
Goodness of fit R2 = 0.85 n.a.


Sources: Gardiner and Theobald 2010; World Development Indicators; World Bank data.
Note: GDP = gross domestic product; n.a. = not applicable.




Short-Run Employment Effects of Infrastructure Investment 35


Although the adjustment by GDP per capita attempts to correct for
differences in labor costs, there may also be substantial differences in the
production structure of economies. Thus, an IO table for one country may
not be representative of the true IO structure of another country. It is also
likely that the larger the difference between per capita GDP levels of two
countries, the greater the difference in the IO structures. More importantly,
resource endowments tend to have a large effect on the structure of pro-
duction even at similar levels of income per capita. So it is expected that
the IO table of a major developing oil exporting country (OEC) would
differ substantially from the IO table of an oil importing country (OIC).


Estimating the Cost of Creating Jobs in
Oil Importing MENA Countries


The availability of non-OECD countries’ IO tables with employment and
multiplier data is limited. Three relevant studies discussing employment
data from an IO table are described in table 3.3. The case of Malaysia is the
least relevant to MENA due to the country’s high income level. South
Africa’s case is more relevant as its income level is closer to that of develop-
ing MENA. Still, both studies rely on outdated IO tables6 that do not con-
tain information on direct employment as a separate category. The study on
Egypt (ILO 2010) is the most pertinent for the estimation undertaken in
this study.7 Egypt’s IO table is the most recent. It has been upgraded to
include 22 sectors based on International Labour Organization (ILO) data
for 2007/08 (ILO 2010). The IO table has other big advantages. Egypt’s
construction sector is disaggregated into five subsectors, corresponding to
areas favored by the fiscal stimulus of 2009 that might be favored in future
infrastructure packages in the region. Information on sector wages has been
integrated with sector employee compensation data to provide an employ-
ment estimate. The calculation of type I and type II multipliers has also
been carried out for all the sectors.8 Finally, the structure of Egypt’s econ-
omy is relatively similar to the structure of other MENA OICs.


Table 3.2 Construction Sector Hourly Wages in the Arab Republic of Egypt, January
2009


US$


Unskilled Semiskilled Skilled


Basic rate 0.88 1.06 1.25
Total rate 1.06 1.25 1.42
Ratio 0.83 0.85 0.88


Source: Gardiner and Theobald 2009.




36 Infrastructure and Employment Creation in the Middle East and North Africa


Bearing this consideration in mind, the study uses Egypt’s IO table and
multipliers to make calculations for the six MENA OICs whose 2009
GDP per capita in US$ is shown in brackets: Djibouti (1,214), Egypt
(2,270), Jordan (4,212), Lebanon (8,175), Morocco (2,811), and Tunisia
(3,792).9 The Gulf Cooperation Council (GCC) and OECs were judged
to be too different in terms of economic structure and GDP per capita
levels to make extrapolations from the Egyptian case reliable.


In addition to the five construction activities, information is available on
two other broad infrastructure sectors: electricity, and transport and com-
munications (table 3.3). These broad sectors include all types of projects
and activities carried out during the year for which the table was con-
structed, apart from construction in electricity. For example, the electricity
sector would include wages paid for operations and maintenance (O&M)
during the year in question. Transport and communications is even
broader and would include construction in the communications sector.


One can estimate the cost of a job in Egypt in 2009 for each of these
sectors based on the employment (type I and type II) multipliers for
2007/08 in the Egyptian IO table. Table 3.4 presents values in US$ for
the cost of creating one direct job, one direct or indirect job (type I), and
one direct, indirect, or induced job (type II). The data for calculating the
costs of a job (direct, type I, or type II) are taken from ILO (2010). Sector
wages are adjusted for nonwage benefits and sector employee


Table 3.3 Sector Coverage Provided by Various Input-Output Tables


Egypt, Arab Rep. South Africa Malaysia


Year of IO table 2007/08 2003 2000
GDP per capita,


2009 US$ 2,270 5,764 7,030
Categories of


employment
identified


Direct, indirect, induced Direct, indirect,
induced


Direct + indirect +
induced, direct +
indirect


Infrastructure
sectors identified


Construction
Buildings
Roads and bridges
Water and sewage
Electricity stations
Other construction
Electricity
Transport and communications


Electricity and gas
Water
Construction
Communication


services
Transport services


Electricity and gas
Water works and


supply
Building and


construction
Transport and


communications


Sources: Bekhet 2011 for Malaysia; International Labour Organization 2010 for Arab Republic of Egypt; Tregenna
2007 for South Africa.
Note: GDP = gross domestic product; IO = input-output.




37


Table 3.4 Cost of Creating a Job in Selected Infrastructure Sectors in the Arab Republic of Egypt, 2009


Share of
employee


compensation (%)


Average annual
total wage in


2007/08 (LE)


Cost of type II
job in 2007/08


(LE)


Cost of direct
job in 2009


(US$)


Cost of type I
job in 2009


(US$)


Cost of type II
job in 2009


(US$) M(I) M(II)


Electricity 20.8 20,562 65,360 21,700 20,472 14,564 1.06 1.49
Construction
Building 8.2 5,838 49,020 17,805 13,092 10,924 1.36 1.63
Roads, bridges 13.3 1,698 11,765 2,858 2,774 2,621 1.03 1.09
Water, sewage 14.7 4,872 27,360 7,376 6,831 6,096 1.08 1.23
Electricity stations 5.2 3,574 51,151 15,387 13,040 11,398 1.18 1.35
Others 23.3 16,548 43,573 15,632 13,247 9,709 1.18 1.61
Transport and


communications 13.5 19,140 78,432 31,808 23,738 17,476 1.34 1.82


Sources: ILO 2010 and World Bank data.
Note: M(I) = type I multiplier; M(II) = type II multiplier.




38 Infrastructure and Employment Creation in the Middle East and North Africa


compensation. Adjustment to 2009 levels was carried out by allowing for
growth in GDP per capita over the period from 2007/08 to 2009 and
using the official exchange rate in 2009 to convert to US$. The most
relevant figures when considering a large-scale program are the costs of
creating type II jobs as these give the upper limit to the short-run number
of jobs created by a given amount of spending, and hence a lower limit
to the cost per job.


The cost of creating a job in the Egyptian construction sector varies
substantially depending on the type of construction activity. The cost of
a new job is the lowest when constructing roads and bridges and the high-
est when constructing buildings. For some sectors, such as roads and
bridges, the costs of a type I job (and of a type II job) are not much lower
than the costs of a direct job. By contrast, in the transport and communi-
cations sector, the cost of a direct job is considerably higher than the cost
of a type I job or a type II job. The differences in the costs of creating
different types of jobs are explained in part by differences in sector wages,
but in all cases these are only a minor fraction of the total costs. The
multipliers are generally quite small, and even in the largest case of trans-
port and communications, the type II multiplier is substantially lower
than that used by Schwartz, Andres, and Dragoiu (2009).


A comparison of these costs with those found in other studies, and
obtained either based on calculations from IO tables or project data,
helps to place these figures in a broader perspective (table 3.5). Such a
comparison is not a straightforward task because other studies give results
for different sectors and time periods, and not all categories of jobs are
presented in every case. Apart from the IO-based estimates from
Tregenna (2007), DIT (2007), Bekhet (2011), and the U.S. studies
quoted by Schwartz, Andres, and Dragoiu (2009), the comparison
includes studies based on World Bank projects in which direct employ-
ment estimates were made. Where possible numbers have been adjusted
by the growth in GDP per capita and converted to dollars to provide
estimates of the cost of jobs in 2009 in US$. These costs per job are used
to calculate the number of jobs that would be generated per billion dol-
lars of spending in the sector—the metric commonly used in studies on
the employment effects of infrastructure investment.


The estimated costs of a job shown in table 3.5 need to be interpreted
with two considerations in mind. First, the cost of one direct job may be
considerably greater than the cost of one type II job, as shown also in
annex 3A and box 3A.1. Second, the costs are expected to vary with the
level of GDP per capita, and several of the countries included are at




39


Table 3.5 Cost of Creating Infrastructure-Related Jobs by Country
2009 US$


Country Source Sector Category of employment Cost per job Jobs per $ billion


Argentina Schwartz et al. Highways Direct 606,060 1,600
Armenia Ishihara & Bennett Lifeline roads improvement Direct 30,400 33,000
Brazil Schwartz et al. Roads Direct 60,325 17,000


Schwartz et al. Rain drainage networks Direct 29,411 34,000
Schwartz et al. Sewerage Direct 45,981 22,000
Schwartz et al. Hydropower Direct 222,222 4,500


Bangladesh World Bank Rural transport improvement Direct 4,412 227,000
Colombia Schwartz et al. Local roads Direct 44,444 23,000


Schwartz et al. Feeder roads Direct 27,907 36,000
Colombia Schwartz et al. Expansion of water and sanitation networks Direct 10,000 100,000
Georgia Vesin International roads Direct 58,824 17,000


Vesin Secondary and local roads Direct 50,000 20,000
Honduras Schwartz et al. Water captation Direct 23,077 43,000


Schwartz et al. Rehabilitation of water networks Direct 17,143 58,000
Schwartz et al. Expansion of water networks Direct 15,000 67,000
Schwartz et al. New treatment plant Direct 40,000 25,000


India DIT ICT—hardware Direct 19,452 51,000
DIT ICT—software Direct 9,214 108,000


Mexico/Guatemala/Perua Vesin Rural road maintenance Direct 2,000–5,000 500,000–200,000


(continued next page)




40


Table 3.5 (continued)


Country Source Sector Category of employment Cost per job Jobs per $ billion


Malaysia Bekhet Electricity and gas Type II 57,381 17,000
Bekhet Water works & supply Type II 27,003 37,000
Bekhet Building and construction Type II 15,829 63,000
Bekhet Transport Type II 20,866 48,000
Bekhet Communications Type II 51,005 20,000


Peru Schwartz et al. Rural electrification Direct 43,478 23,000
South Africa Tregenna Electricity and gas Type II 27,375 37,000


Tregenna Water Type II 31,222 32,000
Tregenna Construction Type II 18,452 54,000
Tregenna Transport Type II 31,036 32,000
Tregenna Communications Type II 32,636 31,000


United States Schwartz et al. Solar PV Direct 370,370 2,700
Schwartz et al. Wind power Direct 294,118 3,400
Schwartz et al. Biomass Direct 1,428,571 700
Schwartz et al. Coal-fired generation Direct 1,333,333 750
Schwartz et al. Natural gas–fired generation Direct 588,235 1,700


Yemen, Rep.a Vesin Rural roads Direct 40,000 25,000


Sources: Tregenna 2007; Bekhet 2011; DIT 2007; World Bank 2010; Vesin 2011; Ishihara and Bennett 2010; Schwartz, Andres, and Dragoiu 2009.
Note: ICT = information and communication technology.
a. Indicates that there was insufficient information to update these figures to a 2009 basis.




Short-Run Employment Effects of Infrastructure Investment 41


Table 3.6 Estimated Costs of a Type II Job in Six MENA OICs, 2009
US$


Country Electricity


Construction


Transport and
communicationsBuilding


Roads
and


bridges


Water
and


sewage
Electricity
stations Others


Djibouti 7,789 5,841 1,402 3,260 6,095 5,193 9,347
Egypt, Arab


Rep. 14,564 10,924 2,621 6,096 11,398 9,709 17,476
Jordan 27,024 20,268 4,865 11,312 21,149 18,015 32,428
Lebanon 52,449 39,338 9,441 21,955 41,048 34,967 62,940
Morocco 18,035 13,526 3,246 7,549 14,114 12,024 21,642
Tunisia 24,329 18,247 4,379 10,185 19,040 16,219 29,195


Source: World Bank data.


levels of GDP per capita substantially above that of Egypt or the MENA
OICs, whereas India, Bangladesh, and Honduras are at the low end of the
range.


The following patterns can also be detected based on the information
in table 3.5. The costs of job creation are strongly linked to the level of
GDP per capita, with the highest costs observed in the United States
and the lowest in Bangladesh and Mexico/Guatemala/Peru. In certain
countries, road programs proved to be low cost, but the nature of the
program appears to influence the cost. Costs are high for national road
construction programs, whereas local and rural road programs are low
cost, and rural road maintenance is least costly. Such low-cost programs
may also be “shovel ready” as explained by Ishihara and Bennett (2010).
Water and sewage programs appear in the midrange of infrastructure
costs, whereas electricity and gas are toward the upper part of the cost
range within a country. These findings are similar to those observed in
Egypt.


The Egyptian data are used to predict type II job costs in selected
MENA countries, using employment multipliers and adjusting for differ-
ences in GDP per capita over time and across countries in 2009. The rela-
tive costs of jobs in the different sectors are the same across countries, but
the general level of job creation costs reflects differences in GDP per
capita. Table 3.6 shows the costs of creating type II jobs in six infrastruc-
ture sectors of the six OICs, and the wide range in the costs of job cre-
ation in the same sector reflects the wide range of GDP per capita found
in these six countries. Table 3.7 shows the costs of creating direct jobs in
the same set of sectors and countries.




42 Infrastructure and Employment Creation in the Middle East and North Africa


Using the estimates in table 3.6, it is possible to derive the number of
type II jobs that would be generated by a given expenditure. This number
is a rough estimate and has limitations. The number of jobs created
depends on availability of suitable labor. Skill shortages will limit the num-
ber of jobs that can be created with a given expenditure. This would need
to be taken into account in designing the size of any investment program.
Even when there are no labor shortages, projects to absorb the investment
spending may not be immediately available. Certain types of projects,
particularly those that simply scale up existing activities, are more likely
to be “shovel ready” than programs focusing on new types of activities.


In cases where there is a need to identify and design projects, there will
inevitably be a delay in obtaining substantial employment effects of infra-
structure spending. Depending on the internal situation of an economy, a
balance may need to be struck between the need to create immediate jobs,
possibly in areas where the output of the job itself is of relatively limited
value, and the need to create strategic jobs that may take longer to plan
and implement. Finally, the calculations reflect adjustments based on 2009
GDP per capita numbers—the latest year for which national accounts
data were available. By 2011, GDP per capita is likely to have changed,
and where it has increased the cost per job would increase and the number
of jobs generated per billion dollars of spending would be lower.


The hybrid approach introduces errors in the estimation of job costs.
The closer the economies’ structure and behavioral parameters to
Egypt’s, the better will be the approximation. The accuracy of the esti-
mates also depends on wage levels being proportional to GDP per capita
differences between Egypt and a given country, and the similarity of the


Table 3.7 Estimated Costs of a Direct Job in Six MENA OICs, 2009
US$


Electricity


Construction


Transport and
communicationsBuilding


Roads
and


bridges


Water
and


sewage
Electricity
stations Others


Djibouti 11,606 9,522 1,528 3,945 8,229 8,360 17,011
Egypt, Arab


Rep. 21,700 17,805 2,858 7,376 15,387 15,632 31,808
Jordan 40,266 33,036 5,302 13,688 28,552 29,006 59,020
Lebanon 78,151 64,120 10,291 26,567 55,414 56,296 114,551
Morocco 26,872 13,526 3,539 9,135 19,054 19,358 39,388
Tunisia 36,251 29,742 4,773 12,324 25,705 26,113 53,135


Source: World Bank data.




Short-Run Employment Effects of Infrastructure Investment 43


mark-up of total wages and benefits over basic wages. The estimation of
sector employment is likely to be most accurate when sectors are defined
the same way across countries. Results will be biased when sector defini-
tions differ from those available in the Egyptian IO table.


Table 3.8 shows the number of type II jobs created per US$1 billion
in each of the six oil importing MENA countries, rounded to the nearest
one thousand. Spending on construction of roads and bridges would gen-
erate more than twice as many jobs as the same amount of spending in
any of the other sectors. Construction of water and sewage infrastructure
is the second most job-intensive activity relative to spending, whereas
transport and communications is the least job-intensive activity.10
Because of per capita income differences, spending of a billion dollars
generates more than six times as many jobs in a sector in Djibouti as in
Lebanon, but the economy of the latter would find it considerably easier
to finance the investment expenditure.


Assuming that each country in the oil importing group has the same
mixture of investment needs, one can use the estimates of annual new
infrastructure needs through 2020 by sector for the oil importing group
(table 2.7) to allocate infrastructure spending across sectors and coun-
tries. With this information, one can then assess the number of jobs cre-
ated by a billion dollar infrastructure investment portfolio by country.
Because the set of sectors distinguished in table 2.6 and the set of sectors
used by the ILO (2010) study on Egypt and presented in table 3.8 are
not identical, it is necessary to assign sectors to the narrower range used
by the ILO (2010). This allocation is shown in table 3.9.


Estimates of the jobs created in response to US$1 billion infrastructure
spending by country and by sector are shown in table 3.10.11 Egypt, for


Table 3.8 Number of Type II Jobs Generated per US$1 Billion of Spending, 2009


Electricity


Construction


Transport and
communicationsBuilding


Roads
and


bridges


Water
and


sewage
Electricity
stations Others


Djibouti 86,000 105,000 654,000 254,000 122,000 120,000 59,000
Egypt, Arab


Rep. 46,000 56,000 350,000 136,000 65,000 64,000 31,000
Jordan 25,000 30,000 189,000 73,000 35,000 34,000 17,000
Lebanon 13,000 16,000 97,000 38,000 18,000 18,000 9,000
Morocco 37,000 73,900 283,000 109,000 52,000 52,000 25,000
Tunisia 28,000 34,000 210,000 81,000 39,000 38,000 19,000


Source: World Bank data.




44 Infrastructure and Employment Creation in the Middle East and North Africa


instance, is expected to create 155,000 jobs for every US$1 billion spent
on infrastructure, whereas the regional average12 is 138,000 jobs. By con-
trast, US$1 billion spending on a “prototypical basket of infrastructure
investment” in the Latin America and Caribbean (LAC) region is esti-
mated to result in just 80,000 type II jobs, according to Schwartz, Andres,
and Dragoiu (2009). The smaller estimate in the LAC region compared
to that of OICs in MENA is expected because the average GDP per
capita in LAC is more than double that of the MENA OICs. For example,
in 2009, LAC’s average per capita GDP was $7,260, whereas that of the
average OIC was US$2,774. The high value of the type II multiplier
assumed by Schwartz, Andres, and Dragoiu (2009) implies that the LAC
region’s job estimate might be somewhat optimistic.


Table 3.9 Shares of Total Investment Needs by Sector for OIC
percent


Sector Share of needs Sector (ILO basis) Share of needs


Paved roads 23.4
Roads 33.1Unpaved roads 9.7


Rail lines 1.8 Other construction 1.8
Ports 1.9 Building 1.9
Telephone mainlines 3.0


Transport and communications 13.0Mobile lines 10.0
Electricity generation 35.3 Electricity stations 35.3
Electricity access 7.2 Electricity 7.2
Water 3.2


Water and sanitation 7.6Sanitation 4.4


Source: World Bank data.
Note: ILO = International Labour Organization; OIC = oil importing country.


Table 3.10 Number of Type II Jobs Created by a US$1 Billion Portfolio
of Infrastructure Spending


Country Population (2009) GDP per capita (2009) Number of type II jobs


Djibouti 864,000 1,214 291,000
Egypt, Arab Rep. 82,999,000 2,270 155,000
Jordan 5,951,000 4,212 84,000
Lebanon 4,223,000 8,175 43,000
Morocco 31,992,000 2,811 126,000
Tunisia 10,432,000 3,792 93,000
OICs average


(population weighted) 136,463,000 2,774 138,000


Sources: WDI; World Bank data.
Note: GDP = gross domestic product; OIC = oil importing country.




Short-Run Employment Effects of Infrastructure Investment 45


The assessment of the employment effects of an infrastructure pro-
gram also needs to consider its duration and nature. A one- or two-year,
stimulus-type program would require projects that can be finished within
the financing timeframe. Such a program would be associated with the
creation of temporary jobs. A more long-term investment program will be
associated with projects that have substantial construction, installation,
and manufacturing (CIM) components (such as renewable energy proj-
ects) and also generate O&M jobs during the life of the new plant. The
labor intensity of O&M is usually different from that of CIM, so it would
be desirable to make a more detailed analysis of the direct job creation of
the various stages of the program, following the approach of REPP
(2001). This requires a detailed specification of the nature of investment
and follow-up surveys with experienced producers.


Alternative Approaches to Estimating the Short-Term
Employment Effects of Infrastructure Spending


The results provided in the previous section are limited to OICs by the
need to rely on the Egyptian IO table for representative calculations. By
following methods different from the one presented in the previous
section, it is possible to come up with estimates of short-term employ-
ment effects from infrastructure spending for all three subregions. For
example, one can adapt the calculation made for the LAC region by
Schwartz, Andres, and Dragoiu (2009) and extend it using calculations
from this study.


Schwartz, Andres, and Dragoiu (2009) investigated the employment
effects of a stimulus package of spending in various infrastructure sectors
in the LAC. They used project data from a number of World Bank studies
to derive the share of investment expenditure on labor, and the share
going to imports, in the main infrastructure sectors. Combining the expen-
diture share on labor with data on region-wide average wages (plus ben-
efits) allowed them to compute the direct employment per US$ billion.
The authors used data from the U.S. highway sector to derive type I and
type II multipliers and calculate indirect and induced employment. Taking
a weighted average portfolio of infrastructure sectors allowed them to
estimate the employment created by a representative basket of infrastruc-
ture investments. Schwartz, Andres, and Dragoiu (2009) estimated that
US$1 billion could generate about 40,000 direct and indirect jobs, or
80,000 induced, indirect, and direct jobs. The ratio of type II to type I jobs
was notably higher than that found in other studies.




46 Infrastructure and Employment Creation in the Middle East and North Africa


A number of extensions and changes were undertaken to adapt the
LAC results to MENA. Wage levels were changed to reflect the fact that
wages are not the same in MENA and LAC, but heterogeneous across and
within MENA countries. Following the idea of Schwartz, Andres, and
Dragoiu (2009), hourly wage costs for construction work are estimated
for the three subregions of MENA as given in table 3.11.13 Computing
hourly wages is crucial for all the estimates and is subject to potential
underestimation, specifically for the GCC and the OECs. To build our
reference values, we have taken data from Gardiner and Theobald14
(2009) for Lebanon, Qatar, and the United Arab Emirates, and data from
Tong (2010) for the the United Arab Emirates.15 We have adapted these
data using the average wage as computed by the ILO’s key indicators of
the labor market (ILO 2011), and we use homogenous wages within
subregions as the investment needs are aggregated values. Our estimates
have to be considered carefully as they do not reflect the intensive use of
migrant workers in the infrastructure and construction sectors and the
high variance of wages in OEC and GCC economies. The share of inputs
used in infrastructure projects relevant to MENA countries, but not avail-
able from the LAC study, were adapted from other sources (see annex
3B),16 whereas the investment needs for infrastructure in the MENA
region are used as estimated in chapter 2.


Finally, the estimation of the multipliers is reconsidered in order to
give a more relevant, but conservative, approximation of potential indi-
rect and induced jobs. The previous section surveys several studies for
type II multipliers, indicating an actual employment multiplier effect
that ranges between 1.2 and 3.5 and is significantly lower than the esti-
mated multiplier of 4 proposed by Schwartz, Andres, and Dragoiu
(2009).17 The adaptation uses the multipliers for direct job creation as
estimated in the previous section, but with a different methodology for
calculating the number of direct jobs. A large variance for the hourly
wages in the infrastructure sector is described in the previous section.


Table 3.11 Estimated Hourly Wages in Infrastructure Works, 2010
US$ per hour


Region Qualified workers Nonqualified workers


GCC 4.5 3
OIC 1.5 1
OEC 3.0 2


Source: World Bank data.
Note: GCC = Gulf Cooperation Council; OEC = developing oil exporting country; OIC = oil importing country.




Short-Run Employment Effects of Infrastructure Investment 47


The large variance reflects actual differences across sectors, but these
may be less relevant for short-term infrastructure projects. Figure 3.1
illustrates these differences, with the figure on the top showing the posi-
tive correlation between sector wages and shares of labor inputs as
reconstructed in the Egyptian IO table, and the figure on the bottom
showing negative correlation between sector wages and estimated
shares of labor inputs adapted from Schwartz, Andres, and Dragoiu
(2009). The former is expected in situations where labor and capital are
substitutes, whereas the latter is expected when capital and labor are
complements.


Table 3.12 summarizes the potential for job creation in the near future
in the three subgroups—the GCC, OECs, and OICs. The estimates are
obtained based on the various methods presented in this study.18 The
direct jobs, computed from input shares, are in column (1) of table 3.12.
Estimates based on multipliers as given in the “Hybrid Approaches to
Estimating the Short-Term Employment Effects of Infrastructure
Investment” section and the method presented in “Alternative Approaches


Figure 3.1 Hourly Wages as a Function of the Share of Labor Inputs in Total Costs


2.0


1.0
H


o
u


rl
y


w
ag


es
, 2


00
9


U
S$


H
o


u
rl


y
w


ag
es


, 2
00


9
U


S$


Share of labor inputs, total inputs


0.5


0 0.05 0.10 0.15 0.20


R2 = 0.4623


R2 = 0.3707


1.5


2.0


1.0


0.5


0 0.2 0.4 0.6 0.8


1.5


Source: World Bank data.
Note: Top figure displays sector data from the Egyptian IO table. Bottom figure displays data adapted for MENA
from Schwartz, Andres, and Dragoiu (2009).




48 Infrastructure and Employment Creation in the Middle East and North Africa


to Estimating the Short-Term Employment Effects of Infrastructure
Spending” section are shown in column (2). Estimates using the hybrid
method of estimating the cost of creating jobs in OICs are given in col-
umn (3). The numbers are calculated based on an investment of US$1
billion, using the infrastructure needs derived in chapter 2.


These results suggest that MENA countries could get a significant
boost to job creation through infrastructure investments. Combining the
estimates of new jobs created per US$1 billion invested in infrastructure
with the estimated investment needs implies total annual job creation
of 2.5 million, representing 1.9 percent of the labor force in MENA
(table 3.13).


However, actual job creation in response to infrastructure investment
programs in the region may not be so large and the following caveats
should be kept in mind when using these results for policy discussions.
These estimates should be considered the upper bounds for what is pos-
sible in the short run as they exclude possible economies of scale and wage
increases associated with large, sustained infrastructure investments; leak-
ages in the form of imports; institutional and governance limitations;
shortages of certain types of labor; and the type of investment project.


Table 3.12 Effect of US$1 Billion of Infrastructure Investment on Job Creation
in MENA


(1) (2) (3)


Direct jobs in
infrastructure


Total (from Alternative
Approaches to Estimating the


Short-Term Employment Effects
of Infrastructure Spending


Section)


Total (from Hybrid
Approaches to


Estimating the Short-
Term Employment


Effects of
Infrastructure


Investment Section)


GCC countries 20,859 26,194 —
OIC 86,566 109,236 138,000
OEC 39,454 48,573 —
Method and


limitations
à la Schwartz


et al. (2009)
combined
with
investment
needs


Direct jobs from column (1) and
multipliers from Hybrid
Approaches to Estimating the
Short-Term Employment
Effects of Infrastructure
Investment section


Multipliers and IO
table from Egypt,
Arab Rep.


Source: World Bank data.
Note: GCC = Gulf Cooperation Council; IO = input-output; MENA = Middle East and North Africa;
OEC = developing oil exporting country; OIC = oil importing country; — = not available.




Short-Run Employment Effects of Infrastructure Investment 49


Certain projects simply scale up existing activities, and new projects may
not have immediate employment impact as there is considerable invest-
ment in planning and design.


The sectoral distribution of investment would also affect the number
of jobs created. There are large differences between sectors’ potential to
create jobs through infrastructure investments. In Egypt, the cost of an
infrastructure job is as low as US$2,621 in the roads and bridge construc-
tion sector, but more than four times higher in the electricity-generating
sector, and nearly seven times higher in the transport and communica-
tions sector. Sectors also differ in their propensity to generate indirect
jobs. It depends on the extent to which the sector requires inputs
from other sectors to produce its output. Based on the Egyptian data, the
ratio of all jobs to the number of direct jobs was as low as 1.09 for con-
struction in roads and bridges, whereas it was 1.82 for transport and
communications.


Implications of Using Labor-Intensive Technologies
in the Maintenance of Unpaved Roads


Often the need to save resources and achieve different policy goals affects
the choice between labor- and capital-intensive technologies in infra-
structure. Clearly, the former is more likely to increase the total amount
of employment generated, and it may also reduce overall costs. This pos-
sibility is discussed in this section in the context of unpaved roads
maintenance.


Table 3.13 Estimated Potential Job Creation in Response to Meeting
Infrastructure Needs in MENA


Infrastructure
needs


(US$ billions)
Direct jobs/
US$ billion


Total jobsa/
US$ billion


Labor force
(thousands)


in 2009


Direct jobs as
a share of the


labor force
(percent)


Total jobs as a
share of the
labor force
(percent)


GCC 15.8 20,859 26,194 16,387 2.01 2.53
OIC 10.3 86,566 109,236 61,598 1.45 1.83
OEC 20.7 39,454 48,573 52,884 1.54 1.90
Total 46.8 2,037,900b 2,544,457b 130,869 1.56 1.94


Source: World Bank data.
Note: GCC = Gulf Cooperation Council; OEC = developing oil exporting country; OIC = oil importing country.
a. Total jobs include direct, indirect, and induced jobs created per US$1 billion in the short run.
b. The estimate of total direct jobs in the last row of the table refers to the jobs created by meeting annual
infrastructure needs. This estimate is obtained by multiplying the estimated infrastructure needs for a particular
group with the corresponding direct jobs estimated per US$1 billion, and then summing up across groups.




50 Infrastructure and Employment Creation in the Middle East and North Africa


Labor-intensive approaches to road maintenance can be a good choice
in countries where job creation is a priority. In addition to the positive
impact on employment creation, labor-intensive technologies can lead to
cost savings relative to equipment-intensive alternatives. Devereux and
Solomon (2006) reported that some labor-intensive programs have pro-
vided up to 30 percent cost savings, and others have identified even up to
50 percent savings.


However, focusing only on reducing the overall investment amount is
probably not the best criterion when considering labor-intensive tech-
nologies. The cost structure of labor-intensive infrastructure provision is
different from the equipment-intensive alternative as the former includes
components like training and development of institutional capacity.
Direct comparisons of labor versus nonlabor costs can therefore be
misleading.


In this study, we assume that labor-intensive technologies for
unpaved roads can realize up to 30 percent in cost savings without
hurting quality (see results discussed by Del Ninno, Subbarao, and
Milazzo 2009; Devereux and Solomon 2006). This implies that unit
costs and depreciation rates are lower under the labor-intensive tech-
nology alternative than the capital-intensive one (table 3.14). The use
of labor-intensive technologies reduces investment needs in the region
on average by 0.3 percent of GDP. With the switch from equipment-
intensive to labor-intensive type of technology, annual investment
needs decrease the most in the GCC and the OECs (about US$2 bil-
lion per year), whereas they have a smaller impact in the OICs (see
table 3.15).


Table 3.14 Road Maintenance and Rehabilitation Program by Type of Technology
U.S. dollars


Equipment-intensive
technology


Labor-intensive
technology


Unit costs
Routine 2,000 1,400
Periodic 10,000 7,000
Rehabilitation 15,000 10,500
Maintenance and rehabilitation annual


cost per km
3,650 2,555


Unit costs for new roads 50,000 50,000
Depreciation rate (%) 7.2 5.1


Source: World Bank data.
Note: km = Kilometers.




Short-Run Employment Effects of Infrastructure Investment 51


Annex 3A Constructing Hybrid Estimates of
Employment Linked to Investment


The standard IO model of an economy links the gross output of a sector
to the final demand for that sector and the intermediate demands made
by other sectors for its output, as follows:


X = A X + F, (3A.1)


where X is a vector of gross outputs of the N sectors of the economy;
F is a vector of final demands for these sectors; and A is the N N matrix
of technical coefficients that indicate how much output from sector i is
directly required to produce one unit of sector j. Gross output is mea-
sured in current money terms, as defined in the IO table.


The gross output is then related to final demand by equation 3A.2,
where the coefficient matrix B measures the total amount of sector i that
is required to be produced to satisfy the direct and indirect demands
produced by one unit increase in the final demand for sector j:


X = (I − A)−1 F B F. (3A.2)


To convert output figures into employment figures, a vector of
employment levels per unit of output is required (usually measured in
full-time equivalent employment units) represented by w. The direct
employment effect of one unit increase in final demand for sector j would
be denoted by wj, and the average cost per direct job would be given by
the inverse of wj. The values of employment per unit of output can be
derived by dividing the share of the compensation paid to employees by
the wage plus benefit level of a full-time employee.


Table 3.15 Investment Needs for Unpaved Road Maintenance by Type of Technology


Technology OIC OEC GCC MENA


% of GDP
Labor intensive 0.5 1.7 1.0 1.1
Equipment intensive 0.6 2.1 1.3 1.4
Difference 0.1 0.4 0.3 0.3
US$ million
Labor intensive 1,764 7,541 7,429 16,734
Equipment intensive 2,246 9,497 9,461 21,204
Difference 482 1,956 2,032 4,470


Source: World Bank data.
Note: GCC = Gulf Cooperation Council; GDP = gross domestic product; MENA = Middle East and North Africa;
OEC = developing oil exporting country; OIC = oil importing country.


ë


ó




52 Infrastructure and Employment Creation in the Middle East and North Africa


Hence, one unit increase in the demand for sector j will generate direct
plus indirect employment, as expressed in equation 3A.3:


E w B .j
i


i ij=∑ (3A.3)
The type I multiplier M(I)j for sector j is then defined by equation 3A.4:


M(I) wi Bij ⁄ w .j
i


j= ∑ (3A.4)
To allow for induced effects from consumption generated by extra


incomes from direct and indirect effects, the IO matrix is expanded to
incorporate a vector of household expenditures on each sector when
income increases by one unit. The calculation of the extended B matrix
is then linked to the augmented A matrix, and the calculation of the type
II multiplier proceeds analogously using the extended B matrix.


The employment and multiplier calculations are initially based on the
year of the IO table. It is assumed that over time, all sector wages (used
to calculate the sector employment coefficients) increase with productiv-
ity and inflation by the same rate as nominal GDP per capita. Denoting
this value by g(0) for the year of the IO table, and by g(1) for the year
required, the employment generated by one unit increase in gross output
is given by equation 3A.5:


Ej 1 = w i
i


∑ g 0( )g 1( ) B .ij( )






(3A.5)


As GDP per capita increases, the direct plus indirect employment
generated by a given expenditure declines. The type I multiplier is unaf-
fected by the rise in GDP per capita because the scaling factors enter
both the numerator and the denominator. The cost per job is obtained by
dividing the size of the unit expenditure by the employment created
according to equation 3A.5 (for direct plus indirect jobs), and this indi-
cates that costs per job will rise as g(1) increases relative to g(0).


A similar correction is required for comparisons across economies
because, even if it can be assumed that the physical IO matrices are similar,
the wage rates would be different reflecting general labor productivity dif-
ferences. Equation 3A.5 can be interpreted as providing an estimate of
employment generated in country C(1) based on an actual IO table from
country C(0). If GDP per capita is higher in country C(1) than in country




Short-Run Employment Effects of Infrastructure Investment 53


C(0), the amount of employment generated by the same expenditure
would be lower in C(1) than in C(0), and the cost per job would be higher.


In cases where there is disaggregate employee compensation by skill
levels, and wage rates for the different skill groups, it is possible to esti-
mate employment generated by skill group in response to investment
spending in a sector. For the case of two skill levels, the employment
levels generated in the base year in sector j by a unit increase in demand
can be denoted w`i and wai. The total number of direct and indirect skill
level ` jobs created will be:


E w Bj
i


i i jα = ∑ α . (3A.6)
To estimate the number of skill level ̀ jobs that would be created in year


1, equation 3A.6 is modified following equation 3A.5. That is, all labor costs
are assumed to increase at the rate of growth of nominal GDP per capita:


Eαj 1( )= wαi
i


∑ g 0( )g 1( )










B .ij (3A.7)


The following example illustrates the adjustments discussed in this
annex. In the case of the Egyptian IO table, data are given on total
employee compensation and jobs created. The latter are calculated based
on the basic wage. To the extent that the total payment to employees is
above the basic wage, the calculation based on basic wages would result in
an overestimation of the number of jobs created, whereas the correct esti-
mate is assumed to be 85 percent of the number shown by ILO (2010).


In the year for which the Egyptian IO table was made, the average cost
of one type II job created by investment spending in the water and sew-
age sector in Egypt was LE 27,359. By 2009, it was estimated that the
cost of a job would have risen to LE 32,680 (US$6,096), reflecting the
increase in per capita GDP.


The cost of creating one direct job is higher than the cost of creating
one type II job, as direct job creation does not factor in the knock-on
effects of investment on indirect and induced employment. It is also
important to recognize that the cost of one direct job as calculated from
an IO table is substantially higher than the employee compensation paid
to a single worker. Additional costs are incurred for buying inputs from
other sectors, leakages through imports, taxes on production, and the
operating surplus of companies. In the case discussed in box 3A.1, direct
labor costs accounted for about 15 percent of total spending.




54 Infrastructure and Employment Creation in the Middle East and North Africa


Box 3A.1


Calculation of the Cost of a Type II Job Using
a Hybrid Approach


The IO table for Egypt for 2007/08 provided information for the water and sewage


sector based on incremental sector expenditure of 1,000 Egyptian pounds (LE):


(1) Direct employee compensationa = LE 147
(2) Direct + indirect employee compensationa = LE 192
(3) Direct + indirect + induced employee


compensationa = LE 243
(4) Sector average annual basic wagea = LE 4,141
(5) Sector average annual total wage = (4)/0.85 = LE 4,872
(6) Direct jobs via basic wagea = 0.036 FTE
(7) Direct + indirect jobs via basic wagea = 0.039 FTE
(8) Direct + indirect + induced jobs via


basic wagea = 0.043 FTE
(9) Type I multipliera = (7)/(6) = 1.08
(10) Type II multipliera = (8)/(6) = 1.21
(11) GDP per capita in current LE in 2007/08 = 10,144
(12) GDP per capita in current LE in 2009 = 12,531
(13) Cost of one job (type II) in 2007/08 = 1,000/(8)/0.85 = LE 27,359
(14) Cost of one direct job in 2007/08 = 1,000/(6)/0.85 = LE 32,680
(15) Estimated cost of one job (type II)


in 2009 = (12) × (13)/(11) = 33,797 LE
(16) Exchange rate to US$ in 2009 = 5.544
(17) GDP per capita in current US$ in 2009 = 2,270
(18) Estimated cost of one job (type II) in


current US$ in 2009 = (15)/(16) = 6,096


To illustrate the use of Egyptian data for extrapolation to another country, the


cases of Djibouti and Jordan are illustrated for type II jobs in the water and


sewage sector.


(19) GDP per capita in current US$ in 2009


in Djibouti = 1,214
(20) GDP per capita in current US$ in 2009


in Jordan = 4,212
(21) Estimated cost of one type II job in


Djibouti in current US$ in 2009 = (19) × (18)/(17) = 3,260
(22) Estimated cost of one type II job in


Jordan in current US$ in 2009 = (20) × (18)/(17) = 11,311


Sources: ILO 2010; World Bank data.
Note: FTE = full-time equivalents; GDP = gross domestic product.
a. Indicates data taken from IO table.




Short-Run Employment Effects of Infrastructure Investment 55


Based on the estimated cost of generating one type II job in Egypt in
2009, and allowing for the difference in GDP per capita, the estimated cost
of a type II job created by investment in the water and sewage sector in
Djibouti in 2009 is US$3,260 and US$11,311 in Jordan. This method can
be extended to all sectors for which there is information in the Egyptian IO
table. The variation in wages in the sector, as indexed by GDP per capita,
result in large differences in the costs of job creation between Djibouti and
Jordan, reflecting the more than threefold difference in GDP per capita.


Annex 3B Estimated Shares of Inputs in
Different Types of Infrastructure


Type of
Infrastructure


Qualified
workers


Nonqualified
workers


Domestic
material


inputs
Imported


inputs Other inputs


Paved roads 0.15 0.06 0.49 0.16 0.14
Roads* 0.03 0.09 0.22 0.63 0.03
Rail lines** 0.13 0.01 0.52 0.24 0.10
Ports 0.10 0.10 0.80 0.00 0.00
Telephone


mainlines 0.15 0.15 0.30 0.24 0.16
Mobile lines*** 0.15 0.15 0.30 0.24 0.16
Electricity


generation 0.10 0.00 0.90 0.00 0.00
Electricity


access 0.14 0.07 0.26 0.53 0.00
Water 0.25 0.25 0.40 0.10 0.00
Sanitation 0.08 0.56 0.32 0.04 0.00


Sources: Schwartz, Andres, and Dragoiu 2009; World Bank data.
Notes: *The ratio for job creation in rail versus road in the US was adapted to the estimates of LAC.
**The percentages for railways were adapted from Heintz, Pollin, and Garrett-Peltier (2009).
***The estimates for telecommunication were adapted from Foreman and Beauvais (1999). They found the
capital/network, sales and marketing, and other input shares in the mobile phone industry to be on average
0.273, 0.408, and 0.319, respectively. We adapted these numbers to provide an approximation of the actual share
of labor inputs.


Annex 3C Potential for Job Creation in the
Three Groups of MENA Countries


Sector


Direct jobs/US$
billion, in the


sector


Share of
investment


(percent)


Direct job/US$
billion, in


infrastructure
Total jobs


(weighted)


GCC countries
Paved roads 47,500 12.00 3,800 4,141
Roads 22,500 28.00 4,200 4,577


(continued next page)




56 Infrastructure and Employment Creation in the Middle East and North Africa


Annex 3C (continued)


Sector


Direct jobs/US$
billion, in the


sector


Share of
investment


(percent)


Direct job/US$
billion, in


infrastructure
Total jobs


(weighted)


Rail lines 83,333 0.00 0 0
Ports 41,667 4.00 1,111 1,789
Telephone


mainlines
104,167 0.80 333 447


Mobile lines 104,167 3.60 1,500 2,010
Electricity


generation
25,000 48.40 8,067 10,891


Electricity access 46,667 1.20 373 556
Water 104,167 0.60 417 504
Sanitation 113,333 1.40 1,058 1,280
Total n.a. 100.00 20,859 26,195
OICs
Paved roads 95,000 23.53 22,353 24,359
Roads 45,000 10.08 4,538 4,945
Rail lines 166,667 1.68 1,148 2,090
Ports 83,333 1.68 1,401 2,255
Telephone


mainlines
208,333 3.03 3,782 5,067


Mobile lines 208,333 9.92 12,395 16,609
Electricity


generation
50,000 35.29 17,647 23,826


Electricity access 93,333 7.23 6,745 10,050
Water 208,333 3.19 6,653 8,050
Sanitation 226,667 4.37 9,905 11,985
Total n.a. 100.00 86,567 109,236
OECs
Paved roads 95.000 25.16 11,952 13,025
Roads 45.000 19.57 4,404 4,799
Rail lines 166.667 0.93 318 579
Ports 83.333 0.00 0 0
Telephone


mainlines
208.333 3.08 1,922 2,576


Mobile lines 208.333 6.90 4,310 5,776
Electricity


generation
50.000 34.58 8,644 11,670


Electricity access 93.333 4.47 2,088 3,110
Water 208.333 2.24 2,330 2,819
Sanitation 226.667 3.08 3,486 4,218
Total n.a. 100.00 39,454 48,572


Sources: Bacon and Kojima 2011; Foreman and Beauvais 1999; Gardiner and Theobald 2009; ILO 2011; Schwartz,
Andres, and Dragoiu 2009; Tong 2010; World Bank data.
Note: GCC = Gulf Cooperation Council; OEC = developing oil exporting country; OIC = oil importing country;
n.a.= not applicable.




Short-Run Employment Effects of Infrastructure Investment 57


Notes


1. In some cases this number is available, so there is no need to do the estimation.


2. In some countries, a fine level of disaggregation is used so that a sector can
easily be identified as being representative of a particular planned future
expenditure. In cases where a coarser level of disaggregation is used, the
multiplier is less accurate.


3. The type II multiplier is the ratio of all jobs created to the number of direct
jobs.


4. Schwartz, Andres, and Dragoiu (2009) showed that wages in the construction
sector in South America varied between US$1.39 and US$4.28 an hour, but
based their calculations on an average wage rate net of benefits of US$2.55
an hour across the region.


5. Data for Qatar and the United Arab Emirates were omitted, as they are
extreme outliers with very high GDP per capita and low construction wages.


6. See IO tables used in Bekhet (2011) for the case of Malaysia and Tregenna
(2007) for the case of South Africa.


7. IO tables are available for a number of countries in the MENA region, but
only the IO table for Egypt has been prepared in a form that provides
employment data.


8. Annex 3A discusses the hybrid method of estimating employment and illus-
trates the method of estimating the cost of type II employment creation in
Djibouti and Jordan, using data from Egypt’s IO table.


9. Lack of current GDP data for the West Bank and Gaza precluded the use of
the hybrid method in that case.


10. It is important to note that the estimate for this sector includes all activities, not
just the construction activities within the transport and communications sector.


11. Since the same sector weights are applied to each country, the total number
of jobs created is proportional to GDP per capita—the scaling factor applied
to every sector within a country.


12. The regional average is a weighted average with weights corresponding to
countries’ population numbers in 2009.


13. Schwartz, Andres, and Dragoiu (2009) assume hourly wages gross of benefits
to be respectively US$6 and US$3 for qualified and nonqualified workers in
LAC.


14. The data in Gardiner and Theobald (2007) for Latin America are consistent
with the ones used in Schwartz, Andres, and Dragoiu (2009), allowing us to
use a similar methodology.


15. Converting to US$ for the hourly wage, the median wage in construction is
approximately US$3.70.




58 Infrastructure and Employment Creation in the Middle East and North Africa


16. The percentages for railways were adapted from Heintz, Pollin, and Garrett-
Peltier (2009). The ratio of job creation in rail versus road in the United
States was adapted to the estimates of LAC. The estimates for telecommu-
nication were adapted from Foreman and Beauvais (1999). They found the
capital/network, sales and marketing, and other input shares in the mobile
phone industry to be on average 0.273, 0.408, and 0.319, respectively. We
adapted these figures to provide an approximation of the actual share of
labor inputs. The estimated input shares are given in annex 3B.


17. The multiplier of 4 means there is one indirect job per one direct job and one
induced job per one direct and one indirect job.


18. Sectoral results are presented in annex 3C.


References


Babcock, M., J. Leatherman, M. Melichar, and E. Landman. 2010. “Economic Impacts
of the Kansas Comprehensive Transportation Program (CTP) Highway
Construction and Maintenance Activities.” Report No. K-TRAN:KSU-10-4,
Kansas State University Transportation Center, Manhattan, Kansas. http://trans-
port.ksu.edu/files/transport/imported/Reports/2010/KSU104_FinalRep.pdf


Bacon, R. and M. Kojima. 2011. Issues in Estimating the Employment Generated
by Energy Sector Activities. Sustainable Energy Department, World Bank,
Washington, DC. http://siteresources.worldbank.org/EXTENERGY2/
Resources/MeasuringEmploymentImpactofEnergySector.pdf.


Bekhet, H. 2011. “Output, Income and Employment Multipliers in Malaysian
Economy: Input-Output Approach.” International Business Research 4 (1):
208–23.


CH2MHILL. 2009. Economic Impact Analysis for the Teanaway Solar Reserve,
Kittitas County, Washington. http://teanawaysolarreserve.com/wp-content/
uploads/2010/05/Teanaway_Economic_Impact_Analysis_10_07_09.pdf.


Del Ninno, C., K. Subbarao, and A. Milazzo. 2009. “How to Make Public Works
Work: A Review of the Experiences.” SP Discussion Paper No. 0905, Social
Protection and Labor, World Bank, Washington, DC.


Devereux, S., and C. Solomon. 2006. “Employment Creation Programmes: The
International Experience.” Issues in Employment and Poverty Discussion
Paper 24, Economic and Labour Market Analysis Department, International
Labour Organization, Geneva.


DIT (Indian Department of Information Technology). 2007. E-Readiness
Assessment Report 2005. New Delhi: DIT. http://www.ncaer.org/Downloads/
Reports/E-Readiness2005.pdf.


Foreman, R., and E. Beauvais. 1999. “Scale Economies in Cellular Telephony: Size
Matters.” Journal of Regulatory Economics 16: 297–306.




Short-Run Employment Effects of Infrastructure Investment 59


Gardiner & Theobald. 2009. International Construction Cost Survey. London:
Gardiner & Theobald. http://www.gardiner.com/assets/files/files/f8d864e655
d f f0adcb60ac1ba0 f0cdb0c4a60b04/2008%20In te rna t i ona l%
20Construction%20Cost%20US%20$%20Version.pdf.


Heintz, J., R. Pollin, and H. Garrett-Peltier. 2009. How Infrastructure Investments
Support the U.S. Economy: Employment, Productivity, and Growth. Amherst,
MA: Political Economy Research Institute, University of Massachusetts.


ILO (International Labour Organization). 2010. Measuring the Impact of the
Egyptian Fiscal Stimulus Package 2008/2009. Cairo: ILO.


———. 2011. Global Employment Trends 2011. Geneva: ILO.


Ishihara, S., and C. Bennett. 2010. “Improving Local Roads and Creating Jobs
Through Rapid Response Projects: Lessons from Armenia Lifeline Roads
Improvement Project.” Roads and Highways Thematic Group, Transport
Notes: TRN-39, World Bank, Washington, DC.


LECG (Law and Economics Consulting Group). 2009. The Economic Benefits
from Investment in Advanced Mobile Infrastructure and Services: the Case of
Thailand. London: LECG. http://www.gsmamobilebroadband
.com/upload/resources/files/LECG_Thailand_report_Final_Oct09.pdf.


Pfeifenberger, J., J. Chang, D. Hou, and K. Madjarov. 2010. Job and Economic
Benefits of Transmission and Wind Generation Investments in the SPP Region.
Cambridge, MA: The Brattle Group. http://www.brattle.com/_documents
/UploadLibrary/Upload859.pdf.


REPP (Renewable Energy Policy Project). 2001. The Work That Goes into
Renewable Energy. Research Report 13, REPP, Washington, DC. http://www
.repp.org/articles/static/1/binaries/LABOR_FINAL_REV.pdf.


Schwartz, J., L. Andres, and G. Dragoiu. 2009. “Crisis in Latin America:
Infrastructure Investment, Employment, and the Expectations of Stimulus.”
Policy Research Working Paper 5009, World Bank, Washington, DC.


Scottish Government. 2011. “Input-output-introduction.” Scottish Government,
http://www.scotland.gov.uk/Topics/Statistics/Browse/Economy/Input-
Output.


Tong, Q. 2010. “Wage Structure in the United Arab Emirates.” Institute for Social
and Economic Research. Working Paper, Zayed University, Abu Dhabi.


Tregenna, F. 2007. The Contribution of Manufacturing and Services Sectors to
Growth and Employment in South Africa. Pretoria, HSRC.


Vesin, V. 2011. “Transport and Employment.” MENA Working Note, World Bank,
Washington, DC.


World Bank. 2010. “Bangladesh—Rural Transport Improvement Project,
Implementation Status Results Report.” World Bank, Washington, DC.






61


C H A P T E R 4


Long-Term Employment Effects
through the Growth Channel


This chapter investigates the contribution of infrastructure investments
to growth, and thus job creation. The idea about this contribution is
based on the two-pronged hypothesis that improved infrastructure leads
to economic growth and that economic growth leads to increased
employment. Quantifying this hypothesis requires first of all an under-
standing of the extent to which infrastructure impacts growth. In a recent
meta-analysis of over 100 studies, Estache and Garsous (2011) found that
the effect depends on three main factors: (1) the specific indicators used
to approximate infrastructure1, (2) the time period analyzed2, and
(3) the level of development of the country in question.3 Taking these
characteristics into account facilitates the generation of a relatively robust
estimate of the average elasticity of growth to infrastructure, which can
be used to estimate the average growth impact of various levels of infra-
structure investment on growth in the Middle East and North Africa
(MENA). Second, one needs to assess the impact of economic growth on
job creation, given by the elasticity of employment to growth. The
elasticity of employment to infrastructure can then be derived from the
elasticity of growth to infrastructure and the elasticity of employment to
infrastructure.




62 Infrastructure and Employment Creation in the Middle East and North Africa


Output Elasticity with Respect to Infrastructure


With Yt as the economic output at time t—measured by the gross domestic
product (GDP) of a country—and Inft as the stock of infrastructure at time
t, the basic definition for the output growth elasticity with respect to infra-
structure (Inf) is given by the following:



Y Y Y)( /


(Inf Inf ) / Inf
tt t


t t t
Inf


1 1


1 −1





− −



. (4.1)


This elasticity measures the percentage point increase in output, given
a percentage point increase in the stock of infrastructure. The elasticity
can be estimated using historical data and used to get a sense of the infra-
structure investment needs under various growth scenarios. Alternatively,
one can use it to assess the extent to which infrastructure commitments
considered by governments will lead to growth.


Since there are no specific growth elasticities with respect to infrastruc-
ture for the MENA region or countries, it is necessary to rely on a survey
of international experience in developing countries to obtain an estimate
of the output elasticity with respect to infrastructure. Although there is
quite a large literature on the topic, relatively few studies cover developing
countries. The basic average elasticity used for the MENA sample, and an
associated confidence interval, draw on the estimation results from
10 studies focusing on developing countries presented in table 4.1. These
are relatively high elasticities when compared to those used for developed
countries—a fact consistent with one of the lessons of the meta-analysis
quoted earlier. However, the variance is large enough to be concerned
about the reliability of the average elasticity derived from this sample.


To assess the robustness of the elasticity, we built a confidence interval
for the mean m of this sample. The empirical mean and the standard error
of this sample are given by x = 0.22 and s = 0.17, respectively. Therefore,
assuming normality, we have in the following:


p −1.96 ≤
0.22− µ


0.17
n −1


≤ 1.96⎛
⎝⎜⎜



⎝⎜⎜
= 0.95. (4.2)


Consequently, there is only a 5 percent chance that m does not fall into
the following interval:


0.115 ≤ m ≤ 0.325. (4.3)




Long-Term Employment Effects through the Growth Channel 63


The boundaries of this interval define reasonably robust lower and upper
bounds for the output elasticity with respect to infrastructure in develop-
ing countries. These bounds will be used in this report to compensate for
the lack of a region-specific estimate.


Employment Elasticity with Respect to Economic Output


The employment elasticity with respect to economic output (GDP)
growth Et is defined as the percentage change in the aggregate level of
employment in the economy given a percentage point increase in economic
output at time t:



EE E


YY Y


(


(


(


(


/


/
.E


t


t


t


t


t


t


1 1


1 1
=





− −


− −
ε (4.4)


This elasticity can be used to get a sense of how much job creation
could be expected to result from various growth scenarios. In contrast to
the growth elasticity with respect to infrastructure, there is no need to
extrapolate the values of this elasticity from international estimates as
estimation for the region and countries can be derived using International
Labour Organization (ILO) data.


The elasticities presented in table 4.2 show that some countries’
employment levels are particularly sensitive to growth, indicating that
they could also be more receptive to policy efforts to promote job cre-
ation in the short run. Table 4.2 also shows aggregate elasticities for the


Table 4.1 Studies Providing Estimates of the Output Elasticity with Respect to
Infrastructure


Authors Estimated Elasticity


Dessus and Herrera 2000 0.13
Gwartney, Helcombe, and Lawson 2006 0.17
Khan and Kumar 1997 (Africa) 0.32
Khan and Kumar 1997 (Asia) 0.26
Nazmi and Ramirez 1997 0.13
Odedokun 1997 0.03
Ram 1986 0.37
Ram 1996 0.30
Ramirez 1998 0.58
Sánchez-Robles 1998 0.00
Sridhar and Sridhar 2004 0.10




64 Infrastructure and Employment Creation in the Middle East and North Africa


region and subregions. These were computed as a weighted average of the
country-specific elasticities, with weights reflecting the share of a coun-
try’s employment in the total regional employment. Although countries
are commonly grouped into regional and subregional aggregates, the
heterogeneity of country-specific elasticities suggests that the practice of
grouping countries might lead to loss of important information. The only
group with some comparability is the one encompassing the OICs as they
all have elasticities below 1.


Employment Elasticity with Respect to Infrastructure


The easiest approximation of the employment elasticity with respect to
infrastructure e can be obtained from the following expression, which


Table 4.2 Employment Elasticities with Respect to GDP in MENA, 2009


Country/region Elasticity


Gulf Cooperation Council (GCC)
Bahrain 0.44
Kuwait 0.41
Oman 0.50
Qatar 1.26
Saudi Arabia 1.00
United Arab Emirates 0.88
Average 0.89
Developing oil exporting countries (OECs)
Algeria 1.29
Iran, Islamic Rep. 0.59
Iraq 0.49
Libya 0.49
Syrian Arab Republic 0.65
Yemen, Rep. 1.12
Average 0.78
Oil importing countries (OICs)
Egypt, Arab Rep. 0.82
Jordan 0.69
Lebanon 0.52
Morocco 0.50
Tunisia 0.55
Average 0.69
MENA 0.77
Middle East 0.71
North Africa 0.84


Source: International Labour Organization (ILO) (key indicators of the labor market).
Note: MENA = Middle East and North Africa.




Long-Term Employment Effects through the Growth Channel 65


combines the employment-growth elasticity and the growth-infrastructure
elasticity:



E E E


YYY


YYY)


)


(


(


/


/


(


(


(
(


/


Inf Inf I/ nf
t


t


t


t


t tt


ttt


t


t


1 1


1 1


1


1 1


1=



×





−−


− −−


−−
ε . (4.5)


The lower and upper bounds for the employment elasticity with
respect to infrastructure can hence be calculated using the latest figures
for employment-growth elasticity and the lower and upper bounds of
the output elasticity with respect to infrastructure. Table 4.3 presents


Table 4.3 Lower and Upper Bounds for the Employment Elasticity with Respect
to Infrastructure


Country/region Lower Upper


GCC
Bahrain 0.05 0.14
Kuwait 0.05 0.13
Oman 0.06 0.16
Qatar 0.14 0.41
Saudi Arabia 0.12 0.33
United Arab Emirates 0.10 0.29
Average 0.10 0.29
OECs
Algeria 0.15 0.42
Iran, Islamic Rep. 0.07 0.19
Iraq 0.06 0.16
Libya 0.06 0.16
Syrian Arab Republic 0.07 0.21
Yemen, Rep. 0.13 0.36
Average 0.09 0.25
OICs
Egypt, Arab Rep. 0.09 0.26
Jordan 0.08 0.22
Lebanon 0.06 0.17
Morocco 0.06 0.16
Tunisia 0.06 0.18
Average 0.08 0.23
Middle East 0.08 0.23
North Africa 0.10 0.27
MENA 0.09 0.25


Source: Based on ILO data.
Note: GCC = Gulf Cooperation Council; MENA = Middle East and North Africa; OEC = developing oil exporting
country; OIC = oil importing country.




66 Infrastructure and Employment Creation in the Middle East and North Africa


lower and upper bounds of the employment elasticity with respect to
infrastructure at the regional and country levels. The information
contained in table 4.3 implies that a 1 percent increase in the infra-
structure capital stock would increase employment in the MENA
region by between 0.10 and 0.25 percent. The employment growth
elasticities and these bounds form the foundation on which simulations
on the impact of infrastructure expansion on jobs in the region can be
undertaken.


Long-Run Employment Response to Infrastructure Investment


This section estimates the employment response induced by infrastructure
investment resulting in 1 percentage point additional growth using
equation 4.4. The wide range of employment-infrastructure elasticities
presented in table 4.3 implies that the infrastructure investment required
to boost growth by a percentage point would vary by country. The lower
the growth elasticity with respect to infrastructure, the higher the
required increase in the stock of infrastructure. For example, the lower
bound of the elasticity suggests that an increase of 8.7 percent in the stock
of infrastructure is required to add a percentage point to growth in the
MENA region. This is the more likely scenario in high-income MENA,
comprising the GCC economies and some upper middle-income MENA
countries, as in the more developed countries the likely growth impact of
an additional unit of infrastructure investment tends to be smaller. In the
case of the upper bound elasticity, the required increase in the infrastruc-
ture stock is just 3.1 percent. However, the estimated employment
response to growth of 1 percent depends only on the employment growth
elasticity and the level of employment in 2009. Thus, we get one set of
estimates by subgroup and these are shown in table 4.4.


The results suggest that the employment response induced by
infrastructure investment, resulting in an additional growth of 1 percentage
point, is expected to lead to a total of slightly more than 9 million addi-
tional jobs in the course of 10 years in MENA, or a little less than
1 million jobs per year. Half of these jobs are expected to be located in
the Middle East and the other half in North Africa. This response is sig-
nificant and accounts for approximately 30 percent of the jobs created
during the 2000s. Had these jobs been created during the 2000s, the
unemployment rate would be substantially lower than the 10 percent
registered at the end of the 2000s. Technological choice (i.e., a switch to
labor-intensive technologies) could enhance the employment creation




Long-Term Employment Effects through the Growth Channel 67


effect of infrastructure investment. Other factors may also make a differ-
ence, such as increased labor mobility, improved procurement rules, and
training and transition subsidies. These estimates depend on the magni-
tude of the employment-growth elasticities in table 4.2. The lower these
elasticities, the lower will be the growth response of additional growth in
response to infrastructure investment.


Notes


1. Using energy as a proxy guarantees a much stronger impact than using water
or even telecoms, and a synthetic indicator provides an intermediate level of
impact as expected.


2. The impact was stronger in the 1950s and 1960s than in the last two decades.


3. The less developed the country, the higher the likely impact. However, this
result is not as statistically robust as expected.


References


Dessus, S., and R. Herrera. 2000. “Public Capital and Growth Revisited: A Panel
Data Assessment.” Economic Development and Cultural Change 48: 407–18.


Estache, A., and G. Garsous. 2011. “Drivers of the Impact of Infrastructure on
Growth: A Meta-Analysis.” ECARES working paper, Université Libre de
Bruxelles, Brussels.


Gwartney, J., R. Holcombe, and R. Lawson. 2006. “Institutions and the Impact of
Investment on Growth.” Kyklos 59 (2): 255–73.


Khan, M., and M. Kumar. 1997. “Public and Private Investment and the Growth
Process in Developing Countries.” Logistics and Transportation Review
59 (1): 69–88.


Table 4.4 Employment Response to Infrastructure Investment Resulting
in a Percentage Point Additional Growth
millions, over a decade


Country/region Number of jobs


GCC 1.4
OEC 4.7
OIC 2.9
Middle East 4.5
North Africa 4.5
MENA 9.0


Source: Based on 2009 employment levels from International Labour Organization (ILO).
Note: GCC = Gulf Cooperation Council; MENA = Middle East and North Africa; OEC = developing oil exporting
country; OIC = oil importing country.




68 Infrastructure and Employment Creation in the Middle East and North Africa


Nazmi, N., and M. Ramirez. 1997. “Public and Private Investment and Economic
Growth in Mexico.” Contemporary Economic Policy 15 (1): 65–75.


Odedokun, M. 1997. “Relative Effects of Public versus Private Investment
Spending on Economic Efficiency and Growth in Developing Countries.”
Applied Economics 29 (10): 1325–36.


Ram, R. 1986. “Government Size and Economic Growth: A New Framework and
Some Evidence from Cross-Section and Time-Series Data.” American
Economic Review 76: 191–203.


———. 1996. “Productivity of Public Capital and Private Investment in
Developing Countries: A Broad International Perspective.” World Development
24: 1373–78.


Ramirez, M. 1998. “Does Public Investment Enhance Productivity Growth in
Mexico? A Cointegration Analysis.” Eastern Economic Journal 24 (1): 63–82.


Sánchez-Robles, B. 1998. “Infrastructure Investment and Growth: Some Empirical
Evidence.” Contemporary Economic Policy 16 (1): 98–109.


Sridhar, K., and V. Sridhar. 2004. “Telecommunications, Infrastructure and
Economic Growth: Evidence from Developing Countries.” Working Paper 14,
National Institute of Public Finance and Policy, New Delhi.




69


C H A P T E R 5


Policy Implications


Infrastructure investment has the potential to create jobs in a short span
of time, while providing a foundation for future growth. This is especially
important in the oil importing countries (OICs), where the infrastructure
gap is the greatest and employment needs are growing. However, it is also
likely to be most difficult in these countries because of strained finances.
Going forward, government decisions on the types of spending to expand
and what to downsize in order to achieve balanced budgets will have
important implications for jobs.


Not all jobs are equal, so investments in infrastructure will need to
be prioritized based on the employment and infrastructure needs of the
country. For example, road and bridge construction projects will have a
direct impact on creation of relatively low-skilled jobs. These types of
projects will be especially effective in addressing job-related concerns in
countries where there is a large pool of relatively unskilled and unem-
ployed people. This is the case in many developing countries in the
Middle East and North Africa (MENA), where the majority of the unem-
ployed do not have tertiary education. By contrast, projects in transport
and communication services have large indirect effects and, therefore, the
ability to create a diverse set of jobs for workers with different skill levels.
These projects will appeal to policy makers in countries where the unem-
ployed have the ability to acquire specialized skills relatively quickly.




70 Infrastructure and Employment Creation in the Middle East and North Africa


This chapter discusses the impact of short-term support for employment
creation through infrastructure spending on long-term employment and
complementary policies needed to ensure beneficial outcomes. Public
works and different types of subsidized employment programs have been
used widely to make it easier for people who cannot find unsubsidized
jobs to find employment and acquire on-the-job skills. These programs
are necessary, for instance, to address structural issues, which will not be
addressed through market forces alone as economies grow bigger.
Subsidized employment programs in infrastructure and construction
have typically been used to create employment opportunities for low-
skilled workers.


The first section of this chapter reviews evidence on the effect of
wage subsidies on the long-term employability of workers and suggests
that training plays an important role in boosting job creation in the long
run. The second section analyzes the types of effective training. The third
section offers ways to bring down training costs by effective targeting,
whereas section four looks at the broader question of whether it is
actually desirable to subsidize job creation in the long run. The final sec-
tion discusses the fiscal costs and benefits of job creation programs.
Prudent planning of infrastructure development and execution of infra-
structure projects will be critical for growth and job creation, as poor
governance and weak institutions are the greatest risks in using infra-
structure as a strategy to enhance employment and growth in MENA
countries.


Subsidized Employment Programs and Job Creation


Boosting short-term job creation in developing MENA economies is desir-
able, particularly in the context of recent political developments.
Subsidized employment programs should be designed to ensure that
there is a positive spillover to long-run employment and employability.
The literature suggests that it is possible and relatively easy—although
potentially costly—to start up or catalyze long-term job creation.
Subsidized jobs help beneficiaries become employed, facilitating their
future employment (Bell and Orr 1994; Richardson 1998). Such an effect,
however, requires competences obtained during employment in the infra-
structure sector to be transferable to jobs in other sectors. Many research-
ers also point to risks of downward pressure on wages. Such risks could
result from a potential increase in the supply of qualified workers with
experience in the infrastructure sector. As infrastructure jobs represent




Policy Implications 71


only 15–25 percent of total jobs, the risk of wage pressures may not be
vast but cannot be ignored.


Subsidized employment programs tend to enhance short-term job
creation but have no impact on future employability (McCord and van
Seventer 2004; Richardson 1998). This is because of (1) stigmatization—
having a subsidized job is a signal of not being able to get a “real” job,
and this can be used by firms as a screening device (Burtless 1985)—and
(2) disqualification—the beneficiaries are often asked to take unskilled
jobs that do not enhance their employability. For instance, Boeri (1997)
has found that participation in public works schemes can reduce the
chance of finding a regular job.


Research suggests that one way of offsetting any negative impact of
subsidized employment programs is to ensure proper training of subsi-
dized workers. When combined with training and counseling, subsidized
programs significantly increase future employability (Katz 1996; Martin
and Grubb 2001). The longer the duration of employment under the
subsidized programs, the higher is the probability of future inclusion in
the “normal” labor market (Gagliarducci 2005). The training provided as
part of the subsidized job addresses underprovision of training by private
firms when the corresponding human capital is transferable.


During the early 1990s, the World Bank conducted research on this
topic, as summarized in Rama (2003). One World Bank project concerning
railways in Brazil provided a unique experience in the creation and design
of training programs, aimed at minimizing the time spent unemployed
and maximizing the workers’ chances of finding a job outside of infra-
structure (Estache, Schmitt, and Sydenstricker 2000). The project
included financing for training of workers who had lost their jobs as a
result of privatization programs and also provided experience in building
national capacity to run decentralized training centers, and in some
instances, the outsourcing of training. The project also provided an ex-post
evaluation of the design of incentives given to workers to enroll for the
training programs.


Types of Training for Lasting Job Creation


Experience shows that the design of training should be given as much
attention as the design of subsidized employment programs. Research
suggests that if there is a role for subsidies, in particular in long-term
programs, those subsidies should be focused on sectors where the market
alone fails to provide sufficient employment, in other words, where there




72 Infrastructure and Employment Creation in the Middle East and North Africa


is a mismatch between the supply and demand for skills. As many workers
are unskilled, the challenge is to identify the various types of training
required.


Ideally, training should start with a general orientation aimed at pro-
viding a basis on which more specialized training can be done as needed.
In theory, workers are more likely to acquire general training when mar-
kets are competitive and the turnover is high (Wasmer 2006). However,
subsidized jobs for low-skilled workers may reduce incentives to become
skilled—an effect amplified by the fact that taxes used to pay for subsi-
dies may result in an additional tax burden for skilled workers (Oskamp
and Snower 2006). Also, there is no reason to provide training specific to
the infrastructure sector if the subsidized job is temporary and the objec-
tive of the training is to facilitate inclusion into the general labor market
rather than the infrastructure sector.


Specialized training is used in particular when severance costs and
labor market frictions are high. To the extent that the time spent in train-
ing is paid, added layers of training may act as an opportunity to buy
transition time until workers are able to get a job. This effect could be
seen during the railway project in Brazil. Workers who took part in the
project were offered two types of training, but most found a job without
completing the more specialized training (Estache, Schmitt, and
Sydenstricker 2000), suggesting that specific training has limited impact
in practice.


The training program for the displaced railway employees in Brazil also
showed that when offered training options covering a wide range of skills,
workers know exactly what they want and how to get trained to improve
their chances of finding a job. In the railway project in Brazil, the majority
of workers self-selected to participate in general training on how to run a
business (Estache, Schmitt, and Sydenstricker 2000). Similar outcomes
were illustrated by Loewenstein and Spletzer (1999) for U.S. workers.
Data from Ireland confirm that general training raises productivity, but
the same cannot be said about specialized training (Estache, Schmitt, and
Sydenstricker 2000).


Overall, the main message of this literature is that if the goal of subsi-
dized jobs is to provide support for inclusion in the job market, the design
of training needs to be part of the design of the job creation program.
Specific training should be considered only if there is market demand for
these skills or if there is a need to buy time in a labor market restructuring
transition. Often, general training supporting labor market flexibility will
be sufficient and more efficient in increasing productivity.




Policy Implications 73


Minimizing the Cost of Job Creation Targeting


Different types of job subsidy targeting strategies involve different types
of implementation and monitoring costs as well as different degrees of
effectiveness (Amin, Das, and Goldstein 2008). There are relatively easy
solutions to reduce these costs, but they take time to be put in place. In
MENA, vouchers may be considered to reduce the costs of targeting as
they are more efficient than direct subsidies, and targeting the long-term
unemployed is more efficient than the less qualified (Brown, Merkl, and
Snower 2011).


The design of targeting practices is essential to the effectiveness of the
program. When targeting is not direct, firms will potentially select benefi-
ciaries that would have been more likely to find a job without the transfer
policy (Marx 2001). This increases the risk of deadweight loss through
substitution effects. There is also a risk that, if the measures have not been
designed to target a sufficiently large range of potential beneficiaries,
some employers will not take the time and energy to use them. Finally,
the real risk is that only large and public firms will benefit from these
programs because these firms have the capacity to mobilize the resources
needed to capture the subsidies. There is thus a tradeoff between the fact
that generous measures generate a greater response but also a greater
burden.1


One way to avoid the risks and costs of direct targeting is to design
subsidies such that workers self-select for the subsidized jobs. The objec-
tive of self-targeting policies is to ensure that certain categories of work-
ers, poor people or women for instance, self-select into the subsidized
jobs, whereas the nontargeted groups choose regular jobs. The subsidies
must therefore be such that targeted workers are willing to accept the
job (participation constraint) and do not have a better job opportunity
(incentive compatibility). Similarly, incentive compatibility must be
such that nontargeted groups refuse the subsidized jobs. However, if the
resulting wage is too low, this self-targeting subsidy can tend precisely to
emphasize the wage gaps and stigmatize a category of workers (Devereux
and Solomon 2006).


Is Subsidizing Job Creation a Sustainable Policy?


When considering the option of subsidizing the creation of jobs, govern-
ments invariably face the challenge of determining the optimal duration
of subsidizing infrastructure. In case of a temporary economic downturn,




74 Infrastructure and Employment Creation in the Middle East and North Africa


short-term wage subsidies are a good option to avoid hysteresis.
A short-term policy of wage subsidies increases the probability of benefi-
ciaries being employed (Bell and Orr 1994; Betcherman, Daysal, and
Pages 2010; Forslund, Johansson, and Lindqvist 2004; Kangasharju 2007;
Katz 1996; Marx 2001; Sianesi 2002). However, there is a deadweight
loss of subsidizing jobs that can be quite significant. The main risk comes
from substitution effects, which can lead to some categories of workers
being priced out of the market by subsidized ones. One of the main chal-
lenges is therefore to ensure that the subsidized jobs are actually “new”
jobs. It has been documented that when firms are required to certify that
they have created new jobs, the deadweight loss can be lower (Marx
2001; Kangasharju 2007).


However, the costs associated with the failure to get it right are high.
Comparing two programs for Turkey, Betcherman, Daysal, and Pages
(2010) found that, for one of the programs, between 47 and 78 percent
of the subsidized jobs would have been created without the subsidy.
In the second one, the efficiency was higher, and they found that between
23 and 44 percent of the subsidized jobs would have been created with-
out the subsidy. The better outcome in the second case can be explained
by the fact that under the second law, the job creation objective was more
specific and firms had to increase their total employment by at least
20 percent to be eligible for subsidies.2


The Turkish result is unfortunately quite representative. An earlier
study argued that in various other countries, leakages of this type were
estimated to be above 66 percent (OECD 1993).3 In the United States,
Bishop and Montgomery (1993) showed that for a very general tax credit
widely used in the 1990s, at least 70 percent of the credits were estimated
to be payments for workers who would have been hired without the
subsidy.4 Martin and Grubb (2001) reviewed various studies and reported
that such leakages were up to 90 percent in Australia, Belgium, Ireland,
and the Netherlands.


In poor countries, infrastructure subsidies bring an additional concern
arising from the potentially perverse incentive that entices people away
from agricultural jobs. A way to reduce such a risk could be to support
infrastructure in the off-season for farming. The impact of short-term
infrastructure subsidies could also lead to competition for workers from
other sectors and an upward pressure on wages, with potentially negative
consequences for competitiveness. The poverty reduction impact of such
higher wages is also questionable, as they may not translate into higher
buying power because the prices of key goods and services consumed by




Policy Implications 75


the poor could increase as well. Targeting of short-term, relatively
specialized, infrastructure jobs may reduce substitution effects. However,
if the wage elasticity is lower in the sector, total job creation from a given
short-term budget may end up being lower than hoped for (Bucher 2010;
Gerfin, Lechner, and Steiger 2005; Sianesi 2002).


Long-term subsidized programs are typically considered in response to
mass layoffs from a major economic restructuring. The main challenge in
these cases is to avoid sustaining sectors or activities, which have no pros-
pects for future development (ILO/IMF 2010). In the case of infrastruc-
ture, subsidized work programs can contribute two types of jobs: (1) those
that support the investment components of the sector (known as CAPEX)
in the short term and (2) long-lasting jobs created to operate and maintain
the long-lived assets (these expenditures are known as OPEX) in the
industry.


When committing to support jobs over a longer period of time, the risk
of generating perverse incentives would potentially increase as there
might be a notion that subsidies would be permanent. One such perverse
incentive is the effect long-term wage subsidies have on displacement
costs, that is, job losses in nonsubsidized firms through distortion of com-
petition (Marx 2001). Also, if job subsidies permit people to enjoy gener-
ous unemployment benefits, people might switch from relying on
benefits to relying on subsidized jobs instead of entering the labor market
(Sianesi 2002), and some categories of workers might be locked in tem-
porary and subsidized jobs (Van Ours 2004).


There are some positive aspects too. Subsidies can compensate for
the implicit tax on severance imposed by employment protection and
avoid displacement costs, if the value of the subsidy is higher than sev-
erance costs (Galasso, Ravallion, and Salvia 2004; Mortensen and
Pissarides 2003). Experience from Finland shows that no displacement
costs were observed because subsidized jobs had to be new and only
one-third of the wage was being subsidized (Kangasharju 2007).
Finally, it is important to keep in mind that long-term programs can be
seen as a redistribution device (Brown, Merkl, and Snower 2011), but
that the odds of generating a lot of perverse incentives in the process
are quite high.


In sum, short-term subsidized work programs can be used more
efficiently than long-term programs to facilitate inclusion in the labor mar-
ket. Wage subsidies in infrastructure works can be designed to limit per-
verse incentives but to do so a serious diagnosis of the local labor market
characteristics is required.




76 Infrastructure and Employment Creation in the Middle East and North Africa


What Are the Net Fiscal Costs and Benefits of Job Creation
Programs?


Infrastructure investments also have macroeconomic implications beyond
questions related to incentives and efficiency. These include the aggregate
fiscal effects of employment subsidy programs and the identification of
added costs and benefits associated with these programs.


One of the biggest benefits of infrastructure employment programs is
the reduction in income transfers needed and the increased formalization
of the economy. Wage subsidies reduce income transfer payments by the
government through increases in employment (Bell and Orr 1994).
Employment creation programs are also often described as a cost- effective
way of providing social protection to the poor (Devereux and Solomon
2006). However, it is not clear how much of a tax benefit such policies
represent if many of the jobs created are low skilled and low taxed.
Ultimately, the dominant effect of subsidies can be to increase social
security registration of firms and workers rather than boosting total
employment and economic activity and taxes, as was found by Betcherman,
Daysal, and Pages (2010) for Turkey and Galasso, Ravallion, and Salvia
(2004) for Argentina.


From a fiscal point of view, there are many potential sources of cost
inefficiency. First, in addition to the direct fiscal cost of providing the
subsidies, as identified in the previous sections, there are also costs like
(1) the deadweight loss of taxation; (2) the cost of demonstration and
training services; and (3) the management costs associated with the costs
of reaching and informing local employers—costs that are usually hidden
in regular government administration. The latter can be significant as the
perceived complexity and administrative costs can be a disincentive for
potential employers to use the schemes. Systems based on vouchers are
usually considered the most efficient, as they significantly reduce the
administrative costs for small businesses (Marx 2001). Some studies
argue that they can be designed as self-financing tools (Brown, Merkl, and
Snower 2011).


Second, costs depend on a number of factors. The first factor is
particularly important for large infrastructure projects and relates to the
mix of locals and expatriates involved in program design and implemen-
tation. The second factor is the choice of delivery mechanism and relates
to the design of procurement rules, that is, the modalities of hiring private
contractors; the wage-rate-setting process; the capital intensity of opera-
tions; the cost of nonlabor inputs; the costs of training provision; and the
administrative capacity and its costs.




Policy Implications 77


Finally, there are more subtle cost drivers that can be quite relevant in
the decision to allocate large fiscal resources to support job creation.
These include the opportunity cost of government spending in the con-
text of limited fiscal capacity; the quality of the institutions, including the
ability to coordinate programs linked to various government levels and
agencies within government; the risk of benefit capture by local elites and
pressure groups; and biased allocation of public resources for political
purposes (Devereux and Solomon 2006).


Recent history shows that in MENA these are not minor risks. The
associated fiscal costs are high in many countries, but they may be minor
compared to the political costs, especially those related to effective coor-
dination between central, regional, and local levels. Such coordination is
vital but problems arise when there are inadequate links between the
different tiers of government.


Overall, it is difficult to predict the short- and long-term fiscal effects
of public infrastructure job programs. They depend on the extent to
which wage subsidies increase the formalization of the labor market,
the impact this formalization has on public revenue and expenditure
obligations, and a large number of indirect and often underestimated
costs associated with these employment support mechanisms. When
considering the real costs of subsidies, one has to take into account the
fiscal capacity of the government, the efficiency of the administrative
process, and the risk of corruption and capture. A recent report on
investment in MENA (World Bank 2011) shows that in economies with
weak rule of law, there is no evidence that public investment stimulates
private investment and growth. In contrast, in countries with an ade-
quate level of property rights’ protection, accountability, and legal insti-
tutions, public investment is strongly linked to growth. In addition,
good rule of law helps attract private investment and countries with
good rule of law show higher levels of investment efficiency.


Concluding Remarks


This study assesses the potential for job creation through infrastructure
investment in the MENA region. The need to achieve tangible employ-
ment results relatively quickly has become an urgent need in the context
of the Arab Spring events. Moreover, heightened regional and global
uncertainty has temporarily restrained private investment, the traditional
source of new jobs in expanding economies.




78 Infrastructure and Employment Creation in the Middle East and North Africa


It is well known that, if effectively directed and fostered, infrastructure
investment has deep and far-reaching impact on economic and social
development. Infrastructure projects can serve as a potential source of
immediate jobs and can boost long-term growth and employment through
associated gains in productivity. The social payoff of developing sustain-
able and integrated basic infrastructure is also significant. Improved provi-
sion of high-quality basic infrastructure services, such as hospitals, schools,
and water supply and sanitation, raises living standards, and enhances
employability of populations and prospects for inclusive growth.


MENA countries have been investing in infrastructure over the
years. Both in the 1990s and 2000s, public investment spending in
MENA was higher than in most developing regions, largely because of
robust spending in the oil exporting countries, which benefited from
rising fuel prices. Spending on infrastructure boosted employment in
the construction sector, which was a major source of job growth in the
2000s relative to other sectors and other countries. Maintaining
momentum in infrastructure spending will be important to keep
growth and job creation from receding.


Although the infrastructure investment in the overall region on the
whole has been strong, there is wide variation across countries in the
quality and quantity of infrastructure. The Gulf Cooperation Council
(GCC) group has the best infrastructure endowments and services in the
region, reflecting high-income levels and commitment to infrastructure
investments financed by oil revenues. However, infrastructure deficien-
cies in developing MENA economies remain a concern. Public invest-
ment spending has been particularly weak in the OICs, which have much
more limited fiscal space than the oil exporting countries. Although
public investment rates increased in the oil exporting countries in the
2000s relative to the 1990s, the opposite happened in the OICs. Recent
growth in public-private partnerships was beginning to fill the gap in
some OICs, but the economic consequences of the Arab uprisings, com-
bined with economic difficulties in Europe, have strained fiscal budgets
in developing MENA economies and reduced private investment, with
possible negative consequences for infrastructure spending. Furthermore,
gaps are likely to magnify as demand for infrastructure grows with popu-
lation and income growth, and as countries tackle challenges related to
water and energy conservation, efficiency, and climate change.


This study estimates MENA’s infrastructure investment and mainte-
nance needs through 2020 at US$106 billion per year or 6.9 percent of
the annual regional gross domestic product (GDP). OECs will need to




Policy Implications 79


commit almost 11 percent of their GDP annually ($48 billion) on
improving and maintaining their national infrastructure endowments,
whereas the OICs and the GCC oil exporters need approximately 6 and
5 percent of their GDP, respectively. Investment and rehabilitation needs
are especially high in the electricity and transport sectors, particularly
roads. Rehabilitation needs will account for slightly more than half of the
total infrastructure needs.


While oil exporting countries will be able to meet their national infra-
structure needs if they maintain investment spending at the rates prevail-
ing in the 2000s, OICs will fall short. As the vast majority of funding for
infrastructure comes from public budgets, it will be critical to protect
public investment budgets and try to increase resources going to the sec-
tor, especially in the case of OICs. Doing so will be a smart choice for
governments looking to create jobs and growth.


MENA’s infrastructure sectors, including construction and infrastruc-
ture services, employ close to one-fifth of the regional workforce or 18.2
million people. About 10.6 million workers are employed in construction,
whereas the remaining 7.6 million provide infrastructure services, but
there are significant variations across countries. Within infrastructure
services, the transport and communication sectors are the biggest employ-
ers, whereas energy and water represent a small fraction of infrastructure
workers.


In addition to being a large employer, infrastructure has the potential
to contribute significantly to employment creation in MENA. In the
short-run, every US$1 billion invested in infrastructure has the potential
of generating, on average, around 110,000 infrastructure-related jobs in
the OICs, 49,000 jobs in the OECs, and 26,000 jobs in the GCC econo-
mies. The region could therefore generate 2.5 million infrastructure-
related jobs just by meeting estimated, annual investment needs, but the
potential varies greatly across countries. Put differently, these jobs would
never materialize if countries instead decide to trim their public invest-
ment rates going forward.


Because of per capita income differences, the spending of US$1 billion
generates more than six times as many jobs in a sector in low-income
Djibouti than in upper middle-income Lebanon, but the latter would
find it easier to finance this investment expenditure. Spending on con-
struction of roads and bridges would generate more jobs as the same
amount of spending in any other infrastructure sector. This is because the
cost of an infrastructure job in the roads and bridge construction sector is
about one-fifth of the cost of a job in the electricity-generating sector, and




80 Infrastructure and Employment Creation in the Middle East and North Africa


slightly less than one-tenth of the cost of a job in the transport and com-
munication services sector. However, sectors differ in their propensity to
generate indirect jobs. It depends on the extent to which the sector
requires inputs from other sectors to produce its output. This indicates
that when investment decisions are made with the objective of creating
jobs, consideration should be given to both direct and indirect employ-
ment effects as well as the type of skills required to implement projects.


The long-term employment effect of infrastructure investment could be
significant. The study finds that the employment response induced by
infrastructure investment resulting in 1 percentage point additional growth
is expected to be 9 million additional jobs in the course of 10 years in
MENA or a little less than 1 million jobs per year. Such a response is sig-
nificant as it accounts for approximately 30 percent of the jobs created in
the region during the 2000s. Had these jobs been created during the last
decade, the unemployment rate would have been substantially lower than
the 10 percent registered in 2009.


A switch to labor-intensive technology could enhance the employment
creation effect of infrastructure investment, and it may also reduce overall
costs. The report discusses the possibility of doing so in the maintenance
of unpaved roads and finds that the use of labor-intensive technology
reduces investment needs in the region by 0.3 percent of GDP. But solely
focusing on costs is probably not the best criterion when considering
labor-intensive technologies. The cost structure of labor-intensive infra-
structure provision is different from equipment-intensive alternatives, as
it includes components like training or development of institutional
capacity. Direct comparisons of labor versus nonlabor costs can therefore
be misleading.


Prudent infrastructure development will be critical for short- and long-
term growth and job creation, as the greatest risk to using infrastructure
as part of an employment and growth strategy in the MENA countries is
poor governance. Not all jobs are equal, so investments in infrastructure
will need to be prioritized based on the employment and infrastructure
needs of the country. For example, road and bridge construction projects
will have a direct impact on the creation of relatively low-skilled jobs.
These types of projects will be especially effective in addressing job-
related concerns in countries where there is a large pool of relatively
unskilled and unemployed nationals. This is the case in most MENA
countries where the majority of the unemployed do not have tertiary
education. By contrast, projects in transport and communication services
have large indirect effects and, therefore, the ability to create a diverse set




Policy Implications 81


of jobs for workers with different skill levels. These projects will appeal
to policy makers in countries where the unemployed have the ability to
acquire specialized skills relatively quickly.


Public works and different types of subsidized employment programs
have been used widely to make it easier for people who cannot find
unsubsidized jobs to find employment and acquire on-the-job skills.
Traditionally, subsidized employment programs in infrastructure and con-
struction have been used to create employment opportunities for low-
skilled workers. But, subsidized employment programs are costly and
should be designed to ensure that there is a positive spillover to long-run
employment and employability. Experience shows that the latter can be
accomplished only if subsidized employment programs are combined
with training and counseling.


Infrastructure investments could provide a quick response and be part
of the solution to MENA’s unemployment challenge, but infrastructure
alone will not resolve this problem. Countries should proceed with
reforms that improve the business environment, especially business regu-
lations and governance. The literature underscores the importance of a
sound regulatory environment and good governance for inclusive growth.
This study focused on estimating the employment impact of infrastruc-
ture investment in MENA. In the future, more work needs to be done to
assess the impact of infrastructure investment on different types of labor,
for example, skilled versus unskilled, young versus old, and domestic
versus migrant workers.


Notes


1. We assume that (1) equal incremental increases in taxes lead to progressively
larger welfare losses and (2) equal incremental increases in each employment
subsidy leads to progressively smaller incremental increases in employment
and social welfare, and a progressively larger government budgetary outlay
(Brown, Merkl, and Snower 2011).


2. The authors compared regions eligible for the subsidies (the “treatment”
group) to regions not eligible (the “control” group). Two types of subsidies
were implemented with different regulations.


3. The authors reviewed a study in Ireland, based on employer interviews, and
Australia, based on interviews with participants in the program—unemployed
workers and employees.


4. This study is based on survey data taken from employers and may therefore
even be underestimating the deadweight loss.




82 Infrastructure and Employment Creation in the Middle East and North Africa


References


Amin, S., J. Das, and M. Goldstein. 2008. Are You Being Served? New Tools for
Measuring Service Delivery. Washington, DC: World Bank.


Bell, S., and L. Orr. 1994. “Is Subsidized Employment Cost Effective for Welfare
Recipients? Experimental Evidence from Seven States.” The Journal of Human
Resources 29 (1): 42–61.


Betcherman, G., N. Daysal, and C. Pages. 2010. “Do Employment Subsidies Work?
Evidence from Regionally Targeted Subsidies in Turkey.” Labour Economics
17 (4): 710–22.


Bishop, J., and M. Montgomery. 1993. “Does the Targeted Jobs Tax Credit Create
Jobs at Subsidized Firms?” Industrial Relations: A Journal of Economy and
Society 32 (3): 289–306.


Boeri, T. 1997. “Labour Market Reforms in Transition Economies.” Oxford Review
of Economic Policy 13 (2): 126–40.


Brown, A., C. Merkl, and D. Snower. 2011. “Comparing the Effectiveness of
Employment Subsidies.” Labour Economics 18 (2): 168–79.


Bucher, A. 2010. “Impacts of Hiring Subsidies Targeted at the Long-Term
Unemployed on the Low-Skilled Labor Market: The French Experience.”
Economic Modelling 27 (2): 553–65.


Burtless, G. 1985. “Are Targeted Wage Subsidies Harmful? Evidence from a Wage
Voucher Experiment.” Industrial and Labor Relations Review 39 (1): 105–14.


Devereux, S., and C. Solomon. 2006. “Employment Creation Programmes: The
International Experience.” Issues in Employment and Poverty Discussion
Paper 24, Economic and Labour Market Analysis Department, International
Labour Organization, Geneva.


Estache, A., J. Schmitt, and E. Sydenstricker. 2000. “Labor Redundancy,
Retraining and Outplacement During Privatization: The Experience of
Brazil’s Federal Railway.” World Bank Institute Working Paper, World Bank,
Washington, DC.


Forslund, A., P. Johansson, and L. Lindqvist. 2004: “Employment Subsidies: A Fast
Lane from Unemployment to Work?” Working Paper 2004:18, Institute for
Labour Policy Evaluation, Uppsala.


Gagliarducci, S. 2005. “The Dynamics of Repeated Temporary Jobs.” Labor
Economics 12 (4): 429–48.


Galasso, E., M. Ravallion, and A. Salvia. 2004. “Assisting the Transition from
Workfare to Work: A Randomized Experiment.” Industrial and Labor Relations
Review 58 (1): Article 6.


Gerfin, M., M. Lechner, and H. Steiger. 2005. “Does Subsidised Temporary
Employment Get the Unemployed Back to Work? An Econometric Analysis
of Two Different Schemes.” Labour Economics 12 (6): 807–35.




Policy Implications 83


ILO/IMF (International Labour Organization/International Monetary Fund).
2010. “The Challenges of Growth, Employment and Social Cohesion,” Joint ILO-
IMF conference in cooperation with the Office of the Prime Minister of
Norway, discussion document, Oslo, Norway, September.


Kangasharju, A. 2007. “Do Wage Subsidies Increase Employment in Subsidized
Firms?” Economica 74 (293): 51–67.


Katz, L. 1996. “Wage Subsidies for the Disadvantaged.” Working Paper 5679,
National Bureau of Economic Research, Cambridge, MA.


Loewenstein, M., and J. Spletzer. 1999. “General and Specific Training: Evidence
and Implications.” The Journal of Human Resources 34 (4): 710–33.


Martin, J., and D. Grubb. 2001. “What Works and for Whom: A Review of OECD
Countries’ Experiences with Active Labour Market Policies.” Swedish
Economic Policy Review 8 (2): 9–56.


Marx, I. 2001. “Job Subsidies and Cuts in Employers’ Social Security Contributions:
The Verdict of Empirical Evaluation Studies.” International Labour Review
140 (1): 69–83.


McCord, A., and D. van Seventer. 2004. “The Economy-Wide Impacts of the
Labour Intensification of Infrastructure Expenditure in South Africa.”
SALDRU/CSSR Working Paper 93, Southern Africa Labour and Development
Research Unit, University of Cape Town, South Africa.


Mortensen, D., and C. Pissarides. 2003. “Taxes, Subsidies and Equilibrium Labor
Market Outcomes.” In Designing Inclusion: Tools to Raise Low-End Pay and
Employment in Private Enterprise, ed. Edmund S. Phelps, 44–73. Cambridge,
U.K.: Cambridge University Press.


OECD (Organisation for Economic Co-operation and Development). 1993.
Employment Outlook. Paris: OECD.


Oskamp, F., and D. Snower. 2006. “The Effect of Low-Wage Subsidies on Skills
and Employment.” Kiel Working Paper 1292, Kiel Institute for the World
Economy, Kiel, Germany.


Rama, M. 2003. “Globalization and Workers in Developing Countries.” Policy
Research Working Paper 2958, World Bank, Washington, DC.


Richardson, J. 1998. “Do Wage Subsidies Enhance Employability? Evidence
from Australian Youth.” Centre for Economic Performance Discussion
Paper 387, London School of Economics and Political Science, London.


Sianesi, B. 2002. “Swedish Active Labour Market Programmes in the 1990s:
Overall Effectiveness and Differential Performance.” Working Paper W02/03,
Institute for Fiscal Studies, London.


Van Ours, J. 2004. “The Locking-In Effect of Subsidized Jobs.” Journal of
Comparative Economics 32 (1): 37–55.




84 Infrastructure and Employment Creation in the Middle East and North Africa


Wasmer, E. 2006. “General versus Specific Skills in Labor Markets with Search
Frictions and Firing Costs.” American Economic Review 96 (3): 811–31.


World Bank. 2011. Economic Developments and Prospects Report, Middle East and
North Africa: Investing for Growth and Jobs. Washington, DC: World Bank.
http://siteresources.worldbank.org/INTMENA/Resources/World_Bank_
MENA_Economic_Developments_Prospects_Sept2011.pdf.






Login