Public Disclosure Authorized

WPS3765

PRIVATIZATION: TRENDS AND RECENT DEVELOPMENTS

Public Disclosure Authorized

Sunita Kikeri and Aishetu Fatima Kolo

World Bank
Public Disclosure Authorized
Public Disclosure Authorized

World Bank Policy Research Working Paper 3765, November 2005

The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the
exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the
presentations are less than fully polished. The papers carry the names of the authors and should be cited
accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the
authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries
they represent. Policy Research Working Papers are available online at http://econ.worldbank.org.
 2

PRIVATIZATION TRENDS AND RECENT DEVELOPMENTS

I. INTRODUCTION

1. This paper takes stock of recent privatization trends, examines the extent to which
government ownership is still prevalent in developing countries, and summarizes emerging
issues in state enterprise restructuring and privatization going forward. It shows that:

ƒ Privatization activity for developing countries as a whole dropped off in the late 1990s
and, while still at overall low levels, is slowly creeping back up towards the more
typical levels prior to the sharp one-time spike in activity in 1997;

ƒ Despite more than 15 years of privatization activity, government ownership is still

prevalent in some regions and countries, and in certain sectors in virtually all regions;

ƒ Despite recent evidence of declining public support for privatization, the evidence
shows privatization’s resounding success in competitive sectors. In infrastructure the
experience is more mixed, with positive outcomes achieved when private participation
is combined with proper market structure, regulatory frameworks, and sound contract
design;.

ƒ Continued efforts need to be made to improve the performance of state-owned

enterprises by: improving privatization policies and institutions; adopting more of a
case-by-case approach for complex sectors and countries; and exposing state enterprises
to market discipline by allowing new private entry and exit of unviable state-owned
firms and improving their corporate governance.

2. The analysis in the paper uses estimated proceeds from privatization transactions that
were carried out between 1990 and 2003 as an indicator of privatization activity. Proceeds are
defined to include all monetary receipts to the government resulting from partial and full
divestitures (via asset sales or sale of shares), concessions, leases, and other arrangements. As
such the data do not cover the following: management contracts; new green field investments;
investments committed by new private operators as part of concession agreements; and the vast
numbers of transfers resulting from mass or “voucher” privatizations across Eastern Europe as
these methods did not generate revenues for government.

3. Data were amalgamated from a variety of sources. For the earlier period, 1990 to 1999,
data are drawn from the World Bank Privatization Database which provides the sale price of a
privatization transaction and the year in which the privatization took place. As the sale price is
recorded on an “announcement” basis rather than on the basis of actual receipts, proceeds do
not necessarily reflect receipts in a particular year since transactions may be paid for over
several years. Data for the more recent years, 2000 to 2003, are aggregated from the following
sources: (i) the World Bank’s Private Participation in Infrastructure (PPI) database; (ii) the
World Bank Africa region’s privatization database; (iii) OECD’s database on privatization in
 3

Africa; (iv) EBRD data on privatization in Europe and Central Asia (the tens of thousands of
enterprises that were privatized through mass privatization or voucher schemes are not covered
in this database); (iv) Privatization Barometer for selected European countries; (v) various other
sources such as Latin Finance and Privatization International; and (vi) various government and
non-government websites. All data are in current U.S. dollars.

4. While proceeds are commonly used to measure privatization trends, several caveats
should be highlighted upfront. First, the heterogeneity of data sources covering a long time
period may inevitably lead to data discrepancies, though every effort has been made to ensure
consistency. Second, proceeds may be underreported in some countries because of lack of data
or poor quality data. Third, data on privatization methods, the percent of shares divested, and
the extent of foreign vs. domestic participation are only available in a few cases or for the
earlier time period, and these factors are therefore not covered in any great detail in the
analysis. Fourth, and related, proceeds from minority share sales may raise revenues but do not
necessarily imply major changes in the control of state-owned enterprises or their improved
efficiency. Fifth, proceeds are subject to swings from a few large transactions in a few
countries and are thus not necessarily indicative of widespread activity or radical restructuring
of the state sector in a particular country or region. Sixth, there are varying definitions of
proceeds but for the most part data are based on gross proceeds rather than net proceeds to the
government, which are likely to be lower after all privatization costs have been covered.
Finally, the analysis includes only central level enterprises, with a few exceptions of state or
provincial level utilities in a few major countries for which such data are available. In light of
these caveats, the set of numbers presented here are meant to be seen as an overall indication of
approximate broad trends rather than an exhaustive listing of each and every transaction in
developing countries.

5. Section II of the paper highlights privatization trends for developing countries as a

whole, followed by regional and sectoral breakdowns. Section III examines the extent to which
government ownership is prevalent in developing countries, by region and sector. Section IV
concludes with a summary of recent developments and issues going forward.

II. PRIVATIZATION TRENDS: 1990 to 2003

6. The reasons for privatization are well established and are not covered in any great detail
here. Developing countries have used privatization as a tool to: improve the productivity of
state enterprises which is typically two to three times lower than private firms and in some
cases significantly lower; access investment capital and improve service delivery of high cost
critical sectors that impact the economy as a whole; and reduce the fiscal burden of loss-
making firms.

7. With these objectives in mind, 120 developing countries carried out 7,860 transactions
between 1990 and 2003, generating close to $410 billion in privatization proceeds, or 0.5
 4

percent1 of total developing country GDP during that period. An overview of the broad trends
over time and by region and sector is provided below.

A. Overview of developing country trends

8. The analysis of overall developing country trends shows that: (i) privatization activity
dropped off after 1997 but picked up, albeit modestly, in recent years; (ii) the average size of a
transaction increased over the years as countries moved towards privatizing larger firms; and
(iii) while a large number of countries are involved in privatization, proceeds are highly
concentrated in a handful of countries.

9. In the early to mid-1990s, privatization proceeds in developing countries averaged

between $20 to 30 billion on an annual basis (Figure 1). Proceeds peaked sharply in 1997 to
almost $70 billion. The sudden and one-time jump resulted from increased activity in large
infrastructure and energy (oil and gas) transactions across virtually all regions, with the largest
share coming from three countries in Latin America (Argentina, Brazil, Mexico), Kazakhstan,
Russia, and China. Revenues declined thereafter as Argentina’s stock of enterprises dwindled
and as activity in Asia and Europe slowed down following the East Asian financial crisis of
1997 and the Russian debt crisis of 1998. By 2001 activity had reached the level of 1990, but
starting in 2002 proceeds began a modest pick up and is slowly creeping back up to pre-1997
levels. The recent increases resulted mainly from share sales in telecoms, power, and banking
in countries such as China (additional share offering of China Telecom), the Czech Republic
(partial sale of Transgas), Slovakia (partial sale of the electricity company), India (telecoms),
Pakistan (United Bank), and Saudi Arabia (telecoms).

1
Over 1990-2003, gross domestic product from the six regions amounted to US$75.8 trillion (in current USD) and US$75.7
trillion (in constant 2000 USD). Using current USD, privatization proceeds as a share of total GDP (from all six developing
country regions) is 0.54%.
 5

Figure 1: Global privatization proceeds

East Asian crisis
Billions of $ 70 Russian debt
crisis

60
Brazil crisis

50
Turkey crisis
40 Mexico
peso Argentina
crisis crisis
30

20 9/11

10

0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

ECA LAC Others

Source: World Bank Privatization Database

10. The average value of transactions increased over time as larger companies were
privatized. While close to 70 percent of the 7,800 plus transactions occurred before 1997, they
consisted mostly of small and medium firms resulting in an average transaction value of $30
million (Figure 2). The bulk of the transactions were concentrated in Europe and Central Asia
where the early years of transition involved a massive transfer of small-scale firms to the
private sector, followed by Latin America and Sub-Saharan Africa. By contrast, fewer but
larger firms were privatized after 1997 in these and other regions (East Asia in particular),
increasing the average transaction size to $96 million. Another indicator of increasing average
size shows the top ten transactions in 1990-99 accounting for 16 percent of total proceeds,
compared to 40 percent in 2000-03.
 6

Figure 2: Number of transactions and proceeds

1,200 70

USD Billions
1,000 60
50
800
40
600
30
400
20
200 10
0 -
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Number of transactions Total proceeds

Source: World Bank Privatization Database

11. While 120 countries have engaged in privatization over the past 14 years, proceeds are
highly concentrated in a handful of countries: over two-thirds of total developing country
proceeds over the entire time period were generated in just ten countries—or 8 percent of all
privatizing countries—with over half of all proceeds generated by the top five alone. While ten
countries consistently generated the bulk of all proceeds, the composition of the group changed
over time (Figure 3). Brazil, Argentina and Mexico dominated the 1990s, with these three
countries alone accounting for virtually 50 percent of all proceeds. Argentina and Mexico fell
off the list in more recent years due to near completion of much of the privatization agenda, but
Brazil remained and together with China, Poland, and the Czech Republic accounted for nearly
60 percent of all proceeds since 2000. For the first time, two countries in the Middle-East and
North Africa region made it to the group of ten on account of the partial sale of Saudi Telecom
and the sale of Regie de Tabac (tobacco manufacturing) in Morocco. Five countries remained
on the list in both periods - Brazil, China, India, Poland, and Russia – representing 41.3 percent
of total proceeds from 1990-2003.
 7

Figure 3: Top ten revenue generating countries

1990-1999 2000-2003

0% 10% 20% 30% 0% 10% 20% 30%

Brazil China

Argentina Brazil

Mexico Poland

China Czech Republic

Hungary Slovak Republic

Poland Saudi Arabia

Malaysia Morocco

India India

Russia Thailand

Peru Russia

Rest of the World Rest of the World

Source: World Bank Privatization Database

12. Data on foreign vs. domestic privatization revenues is available only for the earlier
period, 1990-1999. Foreign revenues comprised 50 percent of total privatization proceeds in
developing countries; three-quarters of foreign revenues were raised through direct investment
and the balance through portfolio flows (World Bank 2001). The Latin American region
represented 56 percent of all foreign investment as a result of sales in telecommunications,
banking, and oil and gas, followed by Europe and Central Asia (23 percent) and East Asia (15
percent). China was the main recipient of foreign investment in East Asia and Pacific followed
by Thailand, Indonesia, and Malaysia. Despite the decline in total privatization proceeds after
their peak 1997 level, foreign investment in privatization remained steady and increased in
1999 due to large transactions in East Asia and Latin America. Data on the FDI share in
privatization is not available for the 2000 to 2003 time period. A comparison of general FDI
trends and privatization trends (Figure 4) shows, however, that while total FDI declined from
2001 onwards, privatization proceeds began to pick up, albeit modestly--largely as a result of
share sales of a few large companies via domestic stock markets.
 8

Figure 4: Total FDI and privatization proceeds from all developing countries, 1990-2003
200
Billions
180
160
140
120
100
80
60
40
20
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

FDI, net inflows Privatization Proceeds

Source: World Bank Development Data Platform and FDI Database

B. Regional trends

13. Proceeds are highly concentrated in a few countries in three regions. Regional shares
have changed over time—with Latin America declining sharply and both Europe/Central Asia
and East Asia increasing their shares in recent years.

Latin America

14. Latin America is the biggest contributor to developing country proceeds, raising $195
billion or 47 percent of total proceeds from 1,300 transactions over the entire time period
(Figure 5). But its share fell dramatically over time from almost 60 percent in the 1990s to
under 20 percent between 2000 and 2003 (Figure 6) and to only two percent at its lowest point
in 2003. The decline is largely attributable to the drop-off in activity in Argentina and Mexico,
which together with Brazil accounted for over 80 percent of regional proceeds in the 1990s,
mainly from telecoms and electricity (65 percent of regional proceeds) and energy (20 percent).
As activity in Argentina and Mexico tapered off due to the dwindling stock of enterprises
and/or political will, Brazil became more or less the only active Latin American country after
2000: large transactions in electricity (CELPE), energy (Petrobras), and banking (Banespa)
accounted for its 85 percent share of regional proceeds, and made it the number two revenue
generator among all developing countries since 2000.
 9

Figure 5: Regional distribution of privatization proceeds: 1990-2003

Billions of USD
40
SA S
M E NA 35
4%
5%
SSA E CA 30
25%
25
3%

EAP
16% 20
15
10
5
-
LA C 1990 1993 1996 1999 2002
47%

ECA LAC EAP

SSA SAS MENA

Figure 6: Regional distribution of privatization proceeds: 1990-1999 and 2000-2003

60%

50%

40%

30%

20%

10%

0%
LAC ECA EAP MENA SAS SSA

1990-1999 2000-2003

Europe and Central Asia

15. Europe and Central Asia—with $101 billion or 25 percent of total proceeds from 4,620
transactions—grew its share over time, raising double the proceeds of Latin America since
2000 and making it the leading revenue earner since then. The early to mid-1990s focused on
small-scale privatization, with large telecoms, power, and oil and gas deals in the mid 1990s
taking place in a relatively small number of countries (Hungary and Russia). During this
period, tens of thousands of enterprises were transferred to the private sector using the small-
scale or “voucher” method or through management-employee buyouts (as these did not involve
any cash transactions or very little in the latter method they are not included in this paper).
These forms of divestiture fundamentally altered the ownership structure of former Soviet
Union economies. In Russia alone, over 15,000 companies were privatized using the mass-
 10

privatization method (Box 1). But the latter part of the decade and the years thereafter saw a
shift in activity to infrastructure sectors: telecoms and electricity transactions now constitute
two-thirds of regional proceeds compared to less than 50 percent in the 1990s. Privatization of
banks is also on the rise, accounting for 10 percent of regional revenues since 2000, with recent
activity taking place in Albania, Bulgaria, Czech Republic, Latvia, Russia, and Slovakia. The
region’s recent rise is attributable to activity in these newer sectors, concentrated largely in
three countries—Poland, the Czech Republic, and Slovakia—which together accounted for
nearly 70 percent of regional proceeds since 2000.

Box 1: Privatization in Russia

Privatization in Russia unfolded rapidly following the collapse in 1991 of the Soviet Union and its
centrally planned economy. In late 1992, about 150 million privatization certificates (vouchers) were
distributed which gave the bearer the right to buy small-scale business or shares at auctions, as well as
to pay for housing. To get the plan through parliament, the reformers agreed to allow managers and
workers to buy 51 percent of shares in businesses, rather than the maximum 40 percent originally
proposed. This often kept enterprises in the hands of "insiders,” Soviet-era bosses with little idea of
how to run private business, creating a delay in the influx of new management and the shake-out of
inefficient companies. The voucher-based scheme ended in mid-1994, marking the beginning of the
second stage of the privatization process, the loans-for-share scheme. By then, 75 percent of small-
scale enterprises had been privatized, along with over 80 percent of the industrial workforce. Overall,
15,000 companies were privatized using vouchers, which accounted for 60 percent of industrial assets.
Cash privatizations or the loans for shares scheme resulted in a compromise where businessmen would
bail out the government with loans and in return would receive shares in big enterprises as collateral.
This was attractive as many crown jewels of Russian industry - oil companies, metal smelters and
mines - had been kept back from voucher privatization. In 1997 the privatization process entered a
third stage, case-by-case privatization. In this phase, financial, insurance, aluminum and coal
companies were sold. By the end of 2001, 129,811 enterprises had been sold, representing about 66
percent of the entire inventory of enterprises at the beginning of privatization. About 700 enterprises
and packets of share were sold in 2001.

More recently international attention has been drawn to the partial re-nationalization of Yukos (the oil
and gas giant) which had been privatized as part of the 1990s program when majority stakes in
lucrative mining and energy enterprises were sold. This occurred when Yukos' main subsidiary,
Yuganskneftegaz, was auctioned off to a little known company which was quickly taken over by a
state-owned firm, Rosneft. The future of Yukos itself is uncertain because of the large tax liability
Russian authorities have imposed on it in the form of back taxes. The president of Russia recently
promised an end to legal challenges to the results of the privatization program, suggesting there would
be no repeat of the Yukos case, which has damaged both investment and economic growth.

Source: Privatization Link.com and Financial Times.com (May 6 2005)

East Asia and the Pacific

16. East Asia and the Pacific—with $66 billion or 16 percent of total proceeds from 420
transactions—also increased its share over time, almost doubling revenues between 1990-99
and 2000-03. One country alone—China—accounted for nearly 90 percent of regional proceeds
in the past four years, compared to 50 percent in the 1990s when Indonesia and Malaysia were
other major contributors (with transactions in transport, electricity, energy and telecoms).
 11

China’s recent share sales in large central enterprises in telecoms and energy made it the top
revenue earner among all developing countries after 2000 (Box 2).

Box 2: Overview of SOE Reform in China

Reform of China’s state-owned enterprises (SOEs) has been a major aim since urban reforms began in
1984. Although there were calls to privatize the SOEs, the Government’s initial emphasis was on
boosting performance by changing the internal governance of SOEs and improving the market
environment in which they operated. By the late 1980s the Government had decided that the best way
to reform small SOEs was to lease them out, with the manager paying the state a fixed proportion of
the firm’s profit. Incorporation was another significant measure that led to privatization.

Since the start of the present century the reform of China’s state enterprise sector has accelerated and
acquired some qualitatively new features. First, the scale of change has expanded to affect almost
every kind of SOE – small, medium, large, and very big; under both central and local control.
Second, ownership diversification has been so extensive that the wholly state-owned non-financial
company has become an endangered species in China. Third, the range of restructuring mechanisms
being used has expanded dramatically to include bankruptcies, liquidations, listings and de-listings,
debt-for-equity swaps, sales to private parties (domestic and foreign), auctioning of state firms and
their assets or liabilities, standard corporate governance techniques, and so on. Finally, mass layoffs –
unheard off just four or five years ago--have become a widespread phenomenon.

Some restructuring of SOEs is occurring through the four state-owned asset management companies
(AMCs) that have been created to take more than $170 billion in nonperforming loans from the big
four state-owned banks. As part of their program, 580 SOEs, accounting for about 40 percent of the
state sector’s assets and sales, have been selected for debt-equity swaps. The AMCs have emerged as
important, and often majority, shareholders in a number of large SOEs.

In the strategically important infrastructure and energy sectors where the regulatory framework is still
evolving, monopolies have been broken and competition has been introduced. Many companies have
been corporatized, and some have been listed on local and international exchanges. China has
nurtured over 20 giant corporations and conglomerates that have proven competitive in the
international market. Some of these companies are laying off tens—or even hundreds—of thousands
of employees, not because they are in financial distress (some of them are hugely profitable) but
because they wish to position themselves as important international players. As of 2002 the top 12
Chinese transnational corporations, mainly SOEs, controlled over $30 billion in foreign assets and
had some 20,000 foreign employees and $33 billion in foreign sales.

Source: Garnaut et al 2005

Middle East and North Africa

17. The Middle East and North Africa region raised $19 billion or 5 percent of total
proceeds from 310 transactions. In the 1990s, activity was concentrated in two countries:
Egypt (with 50 percent of regional proceeds) and Morocco (nearly 40 percent). Transactions in
both countries were mainly in manufacturing, although Morocco’s program was more
diversified with transactions in energy (oil refining) and banking. In 2000, Maroc Telecom was
the first telecoms privatization in the region, raising $1.4 billion. This transaction together with
the partial sale of Jordan telecoms ($508 million) later in that year and the partial sale in 2003
 12

of Saudi Telecom ($4.1 billion) made the telecoms sector the region’s leading revenue
generator in recent years, accounting for nearly 65 percent of regional proceeds since 2000
(compared to less than 1 percent in the 1990s). Among the other recent large transactions in
the region: the sale of Morocco’s Regie de Tabac (tobacco manufacturing) in 2003 for $1.6
billion, and chemicals and cement companies in Egypt.

South Asia

18. South Asia—with 4 percent of total proceeds or $15 billion from nearly 400
transactions—remains at the same levels as in the past. India and Pakistan together account for
75 and 15 percent respectively of South Asian proceeds. Indian revenues were generated
largely from minority share sales in banking and oil and gas, with only a few recent
manufacturing sales transferring strategic control through majority or full share sales and the
divestment of the telecoms company in 2002. Pakistan privatized enterprises in a wide range of
sectors, including telecoms, banking and manufacturing. Sri Lanka had an active program but
its share of regional revenues remained small given the size of its economy. While Bangladesh
recently closed a number of large loss-making jute and textile mills, enterprise sales proceeded
at a slower pace (liquidation is an important aspect that is covered in greater detail in Section
III below).

Sub-Saharan Africa

19. Sub-Saharan Africa—with $11 billion or 3 percent of proceeds and some 960
transactions—had the third highest number of transactions (after Eastern Europe and Latin
America), but 70 percent were mostly small, low-value firms in competitive sectors. While 37
countries were engaged in privatization, the bulk of regional revenues in the 1990s was
accounted for by a few large transactions in Ghana (Ashanti Goldfields and Consolidated
Diamond Mines), South Africa (telecoms, steel, petrochemicals), and Nigeria (selected oil
fields). South Africa was by far the biggest contributor to regional proceeds starting in 2000,
accounting for nearly half of all regional proceeds, mainly due to additional share sales in
Telkom, the sale of South African Airways, and sales in the petrochemicals sector. Other recent
large transactions in the region: the partial divestitures of Mauritius Telecom (for $261 million)
and the Cotton Company of Zimbabwe ($93 million).

C. Sectoral trends

20. Infrastructure (telecoms, electricity/natural gas, transport and water) accounted for half
of all developing country proceeds between 1990 and 2003, followed by competitive sectors
(manufacturing, services, tourism, and other firms), energy (oil and gas, petrochemicals,
hydrocarbons), banking, and the primary sector (Figure 7). While the shares of competitive
sector proceeds declined over time and those of primary and energy sectors remained more or
less steady over time, the share of infrastructure and financial sector proceeds grew in recent
years, accounting together for nearly three quarters of total proceeds compared to 60 percent in
the 1990s (Figure 8).
 13

Figure 7: Sectoral distribution of privatization proceeds: 1990-2003

Energy
15%

Primary
3%

Infrastructure
Competitive 50%
19%

Financial
13%

Figure 8: Sectoral distribution of privatization proceeds: 1990-1999 and 2000-2003

60%

50%

40%

30%

20%

10%

0%
Infrastructure Financial Competitive Primary Energy

1990-1999 2000-2003

Infrastructure
 14

21. Infrastructure proceeds grew over time as more countries started to tackle these sectors
in the late 1990s. They are overwhelmingly concentrated in telecommunications and power,
accounting for 50 and 36 percent respectively of total infrastructure proceeds between 1990 and
2003. Nearly 65 developing countries (or 42 percent of developing countries in the PPI
database) had generated telecoms proceeds through private participation in one form or
another, concentrated largely in Latin America, Europe/Central Asia, East Asia, and more
recently the Middle East and North Africa. In electricity and natural gas, 72 countries (or 46
percent of all developing countries) had some form of private participation in existing utilities,
with activity concentrated largely in Latin America, Europe/Central Asia, and East Asia. The
share of telecom proceeds remained more or less steady over time, while electricity grew
slightly mainly due to recent sales or concessions in Europe/Central Asia. Proceeds in the
transport sector declined from 14 percent in the 1990s (mainly Latin America) to 8 percent
between 2000 and 2003 (in China and Malaysia). And while activity in the water sector grew
from 1 to 3 percent, it still continues to account for a very small share of total infrastructure
proceeds, again mostly concentrated in Latin America (management contracts, which are
excluded from this paper, have been a frequent method of bringing in private sector
participation in the water sector).

22. On the whole, privatization of existing infrastructure assets raised $204 billion since
1990, compared to new green field investments which were significantly higher at $350 billion
(Figure 9).

Figure 9: Total PPI and infrastructure privatization, 1990-2003

Total PPI and Infrastructure Privatization, 1990-2003
Billions of USD
120

100

80

60

40

20

0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Total PPI Infrastructure Privatization Difference

Source: PPI and World Bank databases

Financial sector

23. The share of financial sector proceeds has almost doubled since the 1990s, increasing
from 12 percent to 19 percent of total proceeds in recent years. Banking and other financial
service firms represented 88 percent of financial sector revenues, with insurance firms
accounting for 10 percent and real estate the rest. The increase in financial sector revenues
came mainly from banking transactions in Eastern and Central Europe, driven by EU accession
 15

in countries such as Poland, Turkey, Czech Republic, Bulgaria, Slovakia, and Croatia. Bank
privatizations also took place in Latin America (Mexico and Brazil), East Asia (China,
Philippines, Indonesia and Thailand), South Asia (Pakistan), and sub-Saharan Africa (Nigeria,
Uganda).

Competitive sector

24. Proceeds in competitive sectors, comprising largely manufacturing and service firms,
had the largest number of transactions, registering at over 6,000. Manufacturing firms
accounted for over two-thirds of competitive sector proceeds for the period as a whole, while
service and other sectors accounted for the rest. The largest transactions in manufacturing were
in steel, cement, fertilizer and other chemical sub-sectors. Countries in Europe and Central Asia
were the most active in manufacturing, while Latin American countries divested the largest
share in service and other sub-sectors. Proceeds in the competitive sectors declined from a 22
percent share of total proceeds in 1990-1999 to 9 percent in 2000-2003, reflecting in large part
the shift to activity in infrastructure and finance.

Other sectors

25. Energy sector proceeds (oil and gas, petrochemicals) remained more or less steady:
transactions over the period as a whole were concentrated in a few countries—Argentina,
Brazil, China, South Africa, Kazakhstan, Russia and Poland. The primary sector declined
slightly, with most of the revenues concentrated in Latin America, Europe and Central Asia,
and Sub-Saharan Africa.

III. GOVERNMENT OWNERSHIP: WHERE IS IT AT?

26. The privatization of more, and more important, enterprises would tend to suggest that
government ownership has declined over the years. But a systematic assessment is difficult in
the absence of data on state enterprise sectors, in particular their share of GDP, investment,
domestic credit, formal sector employment, and the like. Such data was last captured for
industrial and developing countries for the period 1978-91 in the World Bank’s Bureaucrats in
Business Report (1995). Since then the 2000 World Development Indicators updated a few
measures for a few selected countries, covering the period 1990-97. But spotty and relatively
old data continue to make a systematic analysis difficult.

27. Nevertheless, anecdotal information shows that while privatization activity has grown, it
appears at the aggregate level to be still small relative to the stock of state enterprises. Since
1990, the average number of transactions per country per year has been only five, compared to
three in the earlier time period from 1980 to 1993 (World Bank 1995), and compared to a much
larger number of enterprises that could have been divested in almost all countries. Moreover,
most countries in fact fell short of the average given the concentration of transactions in a few
countries. This suggests that after more than 20 years of privatization, government ownership is
still prevalent in many developing regions and in certain sectors more or less across all regions
of the world.
 16

A. Regional highlights

Sub-Saharan Africa

28. Government ownership is widespread in Sub-Saharan Africa—in both competitive and

other sectors. While state enterprises in 1991 accounted for 17.3 percent of GDP in African
countries (World Bank 1995), total privatization proceeds since 1990 amounted to only 0.2
percent of regional GDP. Another estimate suggests that state enterprises in Africa still account
for over 15 percent of GDP on average (Chong and Lopez-de-Silanes 2004). In competitive
sectors, over 300 enterprises in over 30 countries that have long been earmarked for
privatization are still pending for sale (OECD 2004); and in most countries in the region, the
water, electricity, telecommunications, railways, airlines, and petroleum product distribution
sectors are still state-owned and operated.

Middle East and North Africa

29. In the Middle East and North Africa region, government’s economic role remains large.
The private sector accounts on average for less than 50 percent of GDP in the region, while the
public sector (including state enterprises) is estimated to account for more than a third of
formal employment compared with 18 percent worldwide excluding China (World Bank 2005).
In countries such as Algeria, Iran, and Syria, state ownership is pervasive: a large share of the
industrial sector is in state hands (80 percent in Syria for example); state-owned banks still
account for 95 percent of banking assets (World Bank 2003); and many of the countries derive
a large share of government revenues from state enterprises (Figure 10). And in the Gulf States,
most of the larger, non-oil industries are government owned (World Bank, 2003). In many if
not most countries in the region, critical services (in transport, power, and finance) are still
provided mostly by the public sector.
 17

Figure 10: Share of government revenues from public enterprises*

% of total revenues

0% 20% 40% 60% 80%

Algeria 60%
Egypt 12%
Jordan 14%
Lebanon 17%
Morocco 4%
Syria 24%
Tunisia 7%
Turkey 6%
Czech Rep. 2%
Hungary 3%
Poland 4%
Slovakia 5%
Slovenia 3%
China 7%
Thailand 7%

* Includes mostly revenues from property and tax payments as a percentage of total revenue. Data refer to the
most recent year available during the period specified.
Source: ANIMA- Euromediterranean Network of Investment Promotion Agencies

Asia -- China and India

30. In China, the state’s economic role declined dramatically from 80 percent of GDP in
1978 to 17 percent in 2003, but state enterprises still remain a dominant economic force,
employing half of China’s 750 million workers, controlling 57 percent of its industrial assets,
and dominating key industries such as financial services, power, and telecommunications
(McKinsey Quarterly October 2004). Only recently has China started restructuring its medium
and larger state enterprises, but largely through minority share sales or new share offerings via
the capital market which came to be viewed as a new source of funds for state-owned
companies given the growing strains in the banking sector. Government ownership still remains
dominant in these firms: non-tradable state shares account for about two-thirds of equity in
listed state-owned companies (Financial Times, March 29, 2005).

31. Similarly, the Indian government launched a “disinvestment” policy in 1991 when
aggregate losses of non-financial public enterprises added to the overall fiscal deficit by about 2
to 3 percent of GDP. In the 1990s disinvestment mainly involved the sale of minority shares to
raise revenues. Starting in the late 1990s and early 2000s, a few centrally owned enterprises
were fully or majority divested to private investors, but these were small compared to the value
of the assets of the public sector. The momentum picked up in 2002 with the sale of telecoms
shares to a strategic partner, but the program came to a standstill later that year when the
proposed sale of two oil companies was postponed. A recent analysis shows that between 1991
and 2004, only 28 percent of planned privatizations were achieved, with 2001-02 being the
 18

most successful year in which 47 percent of privatization targets were met (Oxford Analytica,
March 29, 2005). Meanwhile more than 40 percent of the country’s capital stock is still state-
owned (McKinsey India study 2001). At the sub-national level as well there are a large number
of enterprises in sectors such as manufacturing, agro-processing, power, and transport which
continue to remain in state hands.

Europe and Central Asia

32. State enterprises are playing substantially less of a role in Europe and Central Asia
compared to the 1990s. By 2003, all but two countries in the region had completed or were on
the verge of completing small-scale privatization, while 20 countries had privatized at least 25
percent of large enterprises. The private sector’s share of GDP by 2003 grew to more than 50
percent in 22 countries, compared to only nine in 1994, with eleven of the 22 countries at 70
percent or more (World Bank 2004).

33. Yet, many countries, including EU accession countries with longstanding privatization
programs, e.g. Poland, still have large “strategic” companies in competitive sectors such as
steel, petrochemicals, and manufacturing that are state owned. State ownership is especially
pervasive in Central Asia: in 2002 state enterprises accounted for over 50 percent of GDP in
Belarus, Moldova, Tajikistan, Turkmenistan, and Uzbekistan (World Bank 2004). In the
Balkans, countries such as Kosovo and Serbia have thousands of “socially-owned” firms.
Residual government ownership in privatized enterprises in the region as a whole is also quite
high (Lieberman 2003). And across the region, despite recent advances in sectors such as
telecoms, there is still a large stock of utilities (power and water in particular), banks, and non-
bank financial institutions such as insurance companies that are still state-owned.

Latin America

34. Privatization has been big business in Latin America. But the activity has been
concentrated in the three largest countries-- Argentina, Mexico, Brazil--and a handful of
smaller ones such as Bolivia and Peru. Indeed, these countries have substantially reduced the
role of state enterprises: in Mexico, for example, the share of public enterprises in GDP went
from nearly 15 percent in 1982 to less than 5 percent in 2001 (Box 3). But in most other
countries in the region, state enterprises are still in operation.
 19

Box 3: SOEs in Mexico

The role of SOEs in the Mexican economy has dropped dramatically over the last two decades. In 1982,
there were 1,155 SOEs that accounted for 4.4 percent of the labor force and 14 percent of GDP. By 1990,
the number of SOEs had dropped to 418 accounting for 3.7 percent of employment and 10.1 percent of
GDP. These figures decreased even further during the 1990s when the government continued to withdraw
from many economic activities, including in strategic and priority areas. The privatization process
included ports, several airports, and railroads. Progress was also made in allowing private participation in
secondary petrochemicals and electric power generation.
SOEs share in GDP and employment Number of SOEs

15% 1250
1000
10%
750
5% 500
250
0% 0
1982 1990 2001 1982 1990 2001

GDP Employment Running Divestment in process

In 2001 there were 205 SOEs accounting for 3 percent of GDP and 1.5 percent of employment. In this
year 20.3 percent of the federal budget was assigned to subsidies and transfers of which 31.4 percent was
accounted for by SOEs. Three major SOEs in the energy sector accounted for 2.9 percent of GDP, 0.62
percent of employment, and contributed about 55.9 percent of the federal budgetary income. At the end of
2003, about 80 of the 210 remaining SOEs performed commercial activities in the following areas:
petroleum, electricity, transport, communications, postal services and the commercialization of basic or
mass consumption products. The rest were financial and social institutions with programs in health,
housing, education, culture, food and other products and services. SOEs still maintain control over
important industries such as petroleum (Pemex) and electricity (CFE and LFC) that remain reserved areas
for the State.

Source: Regulating Market Activities by Public Sector, 01-February-2005, Organization for Economic
Co-operation and Development

B. Sectoral highlights

35. Government ownership in existing infrastructure and financial sector entities continues
to exist despite increasing emphasis on private sector participation over the past decade.

Infrastructure

36. Power utilities in nearly 85 developing countries--or 55 percent of all developing

countries in the PPI database--are still owned and operated by the state. In a survey of 52
developing countries with generating capacity of between 29 megawatts (the Gambia) and 318
gigawatts (China), almost 70 percent had not started or completed the process of brining private
sector participation into the sector, and a further 18 percent had just begun the process; by
contrast, independent power providers had been established in 67 percent of the countries with
 20

another 21 percent planning to open electricity markets to them (World Bank 2004). State
ownership is most prevalent in low-income countries with weak institutional and regulatory
capacities. Of the 47 Sub-Saharan African countries, for example, power utilities in 32 remain
fully state-owned and operated, while only eight countries have had concession contracts and
another seven have management or lease contracts (Gökgür 2004). But state ownership is not
prevalent in Africa alone: a recent survey of nearly 20 low-income countries in different
regions--including 11 in South Asia, East Asia, and Europe and Central Asia--shows that while
many have corporatized their state-owned firms and/or brought in the private sector through
IPPs, only one country (Mali) had private participation in distribution (while four are in
advanced stages of the process), and only four countries had private participation in generation
(Briceño-Garmendia et al 2004).

37. Similarly telecoms utilities are still owned and operated by the state in 60 percent of the
developing countries included in the PPI database. State-ownership exists in virtually all
regions. While Sub-Saharan Africa has had some successes with PPI (Mauritania, Uganda for
example), the incumbent operator is still fully state-owned in 27 countries, seventeen of which
have not initiated reform while the remaining ten are either stalled or delayed for political
and/or market reasons (World Bank 2005). The remaining 20 countries in the region have had
varying levels of private participation: one (Somalia) is fully private, 13 have majority private
participation, and six involve minority share sales. In Latin America, telecom operators in 40
percent of the countries are still fully state-owned and operated, while in the Middle East,
South Asia, and Europe and Central Asia the numbers are at 65, 60 and 30 percent respectively
(ITU 2003).

38. The numbers for transport and water are substantially higher, with nearly 60 and 70
percent of developing countries respectively not having any form of private participation in
these sectors. In transport, concessions have been concentrated in Latin America and East Asia,
with relatively minimal activity in other regions.

Financial sector

39. Privatization has decreased government ownership in the banking sector: in Europe and
Central Asia, for example, the portion of total assets in state-owned banks declined from almost
half in 1995 to less than 20 percent in 2003, with the biggest gain in EU accession countries
compared to CIS countries; in Bulgaria, for example, 85 percent of the banking system has
been privatized (Lieberman 2003).

40. But government ownership in developing countries has not been reduced to negligible
levels: in South Asia, India--which represents half of the region’s activity--still has over 70
percent of total assets in public commercial banks, while in the Middle East and North Africa
region as a whole 40 percent of assets remain in state-owned banks (Figure 11). Many
developing countries in other regions still have at least 30 percent share of banking assets under
state control. Among them: Albania, Argentina, Belarus, Brazil, Costa Rica, Romania, Russia,
Thailand, Turkmenistan, and Uruguay (Clarke, Cull and Shirley 2003). In China, state banks
remain a major liability: nonperforming loans of major financial institutions at the end of 2003
were equivalent to about 18 percent of the loans of the banks and 21 percent of GDP, with as
much as 90 percent of loans regarded as government contingent liability (Caprio et al 2004).
 21

Figure 11: State-owned banks are still holding on especially in India and MENA

100
Percentage of bank assets at
private commercial banks

80

60

40

20

0
India MENA EAP ECA LAC Developed

1985 1995 2003

Source: World Development Report 2005

Energy

41. In energy, political issues surrounding private and in particular foreign investment are
paramount in sectors such as oil and gas and thus many such companies remain either fully or
majority state-owned. Some inroads are being made: in Mexico, for example, the need to inject
funds into heavily indebted Pemex, the state-owned oil company accounting for 36 percent of
government revenues, has recently led policy-makers to consider a “new model” for the
company, including the removal of an explicit prohibition on private investment and possible
share offering of shares on the domestic stock market.

42. In short, the above analysis suggests that despite a long track record of privatization over
more than a decade, government ownership still appears to be prevalent in many regions and
sectors.

43. The costs of not reforming state enterprises are high: the productivity of state enterprises
is typically two to three times lower than private firms, and in some cases significantly lower;
they continue to remain a fiscal burden; lack of investment capital has led to poor service
delivery in critical sectors that impact the economy as a whole. A high percentage of sub-
Saharan African and South Asian firms identify poor infrastructure services in
telecommunications and electricity as major or very severe obstacles to private sector
development and growth (World Bank Investment Climate Database). Thus the longer
governments wait to address these issues, the greater the economic and financial burden and the
greater the barriers to achieving economic growth.

IV. EMERGING ISSUES

44. A vast body of empirical work on the impact of privatization that has been summarized
in detail elsewhere shows that in competitive sectors privatization has improved firm
performance (Megginson and Netter 2003 and Kikeri and Nellis 2004 provide recent reviews).
Available studies conclude that privatization in these sectors improves performance and
 22

increases returns to new owners and shareholders, with more robust findings in high and
middle income countries than in low income countries.

45. In infrastructure sectors the evidence is more mixed: recent reviews show that private
participation yields benefits when accompanied by proper policy and regulatory frameworks
(Estache 2004, Harris 2004). Efficiency gains tend to be concentrated in certain regions (Latin
America, East Asia) and sectors (telecoms and electricity). The gains also tend to be
concentrated in urban rather than rural areas; urban areas for example constitute less than 30
percent of the populace in low and middle income countries and have access rates at about 70
percent higher than their rural counterparts (Briceño-Garmendia et al 2005). In general the gains
have been harder to achieve and sustain in low-income, institutionally weak countries and in
sectors such as water where the social implications of tariff adjustment make private
participation more of a challenge.

46. Contract renegotiations are frequent. In Latin America, for example, half of all
infrastructure concession contracts have been renegotiated, with transport and water having the
highest rates (Guasch 2004). Renegotiations are not always bad, and some may be required to
deal with poorly designed contracts, but cancellation or distress as a result of failed
renegotiations is on the rise, totaling 140 or 9.7 percent of all infrastructure projects by the end
of 2003 (Izaguirre 2004). Water and sewerage sectors have been most affected; in Buenos
Aires and Manila, for example, water concessions encountered difficulties when devaluations
led to large tariff increases that were politically difficult to implement. In sub-Saharan Africa,
difficulties in negotiating and enforcing credible contracts have led concessions in the water to
come under stress, and in some cases (e.g. Uganda, Mozambique) to be terminated.

47. Compared to the large number of privatization transactions there have been few
instances of outright reversal or re-nationalization. Those that have taken place usually
occurred when very large heavily indebted privatized firms failed, or in times of crisis and/or
political change. Among the prominent examples: Rail Track in the UK (as a result of
bankruptcy), Air New Zealand (a $300 million rescue package), La Paz and El Alto water
concessions in Bolivia, and banks in Chile and Mexico that had been poorly privatized in the
first rounds with protection from foreign competition, then renationalized, and subsequently re-
privatized.

48. Though relatively few in number these and other recent transactions under stress have
contributed to declining public support, worldwide, for privatization. Recent public opinion
polls show the following:

ƒ In Latin America, about 75 percent of the population in 1995 supported privatization

(Estache 2004), but a more recent Latinobarómetro poll found that between 1998 and
2002 the percentage of people who say that the state should leave productive activity in
private hands fell from 51 to 35 percent. During the same period the percentage of
those who believed that privatization has been beneficial for their country fell from 46
to 28 percent (Figure 12). And when asked specifically whether the state or the private
sector should be responsible for service delivery in the water and power sectors, 70
percent of respondents wanted the state to take charge;
 23

Figure 12: Belief that privatization has been beneficial in their country (%)

Argentina

Uruguay

Paraguay

Chile

Colombia

Bolivia

Mexico

Guatemala
Nicaragua

Panama

Peru

Costa Rica
Honduras

El Salvador

Brazil

Venezuela

Ecuador

0 10 20 30 40 50
Perce nt

Source: Latinobarómetro, 2002

ƒ In Sub-Saharan Africa, a recent AfroBarometer survey of 15 major countries

undertaking economic reforms between 1999 and 2001 found that less than 35 percent
of the population preferred privatization to state ownership -- ranging from 30 percent
or less in Ghana, Lesotho, Mali, Uganda and Zambia to 40 percent or more in
Botswana, South Africa, and Tanzania (AfroBarometer 1999/2001). In Ghana, for
example, 70 percent of respondents opposed the privatization of public enterprises,
with 50 percent “strongly” opposed. In Zambia, focus group researchers found people
to be “deeply attached to public ownership and the public provision of employment”
(Slattery 2004);

ƒ In Eastern and Central Europe, a survey of 16 countries (excluding Central Asian

states) found that 52 percent of the population between 1998 and 2001 supported
private ownership over state ownership, with support being the highest (over 80
percent) in the Baltic states (New Democracies Barometer 1998, 2000). Yet, in Russia,
70 percent of people surveyed felt they were worse off after privatization and preferred
state ownership, while in countries such as Belarus, Moldova and Ukraine less than 40
percent of the population favored privatization;

ƒ In East Asia, a survey of eight countries (excluding Vietnam, Laos and Cambodia)
found that only a third of the population surveyed favored privatization over state
ownership (Asia Barometer 2001-02). While support was quite high in Korea (60
percent), it was low in China (17 percent), Thailand (33 percent), and the Philippines
(39 percent).
 24

ƒ In South Asia, regional polls are not available, but a recent survey in Sri Lanka found
that between 60 and 80 percent of people felt their living conditions worsened after
privatization.

49. Declining public support is fueled by several factors: long traditions of state ownership
in many countries and regions; the overselling of privatization and creation of unrealistic
expectations; generally worsening economic conditions and painful liberalizing reforms that
became associated with privatization; lack of public information; and badly designed
privatization processes leading to poor outcomes and controversies about prices, particularly in
sectors such as water and power (see Harris 2004 for a summary of evidence in these sectors).

50. At the same time, the costs of not reforming state enterprises in critical sectors such as
infrastructure and finance are high. What then needs to be done? It is beyond the scope of this
paper to answer the question in any great detail. It thus only highlights a broad framework for
moving ahead on state enterprise reform, including the need to: (i) keep improving privatization
policies and processes; (ii) bring market discipline to state owned firms; and (iii) pay greater
attention to results measurement.

Continue to improve privatization policies and processes

51. The experience from the past several years shows that privatization is neither a panacea
nor a universal solution that can be easily applied to all countries and sectors. There is no “one
size fits all” approach. Instead, policies and approaches need to vary taking into account
sectoral and country circumstances, in particular market structures and levels of institutional
development.

Competitive sectors

52. In competitive sectors there is no longer much of a debate about privatization or its
impacts: many developing countries have already accomplished privatization or made
significant strides in competitive sectors, while in others such firms can and should be
privatized quickly. Concerns about privatization typically focus on the pricing of assets,
transparency, and labor impacts. The experience shows these concerns can be addressed
through proper valuation techniques, the adoption and enforcement of competitive bidding
processes, setting up of the right institutional and decision-making arrangements, and the
development of labor programs and broader social safety nets for workers negatively affected
by privatization. As discussed in greater detail below, infrastructure and financial sectors are
more difficult by contrast, requiring far greater attention to the development of competition
policies and regulatory frameworks.

Infrastructure sectors

53. Infrastructure sectors--involving complex issues of price, access, and quality of services-
-call for more of a case-by-case approach taking into account prevailing market structures and
regulatory capacities. Particularly in network sectors such as water and power and in low-
income countries the challenges are greater. Care needs to be taken that private participation in
 25

infrastructure is accompanied by policies and institutions that promote broader sector-wide

reform by introducing competition, addressing tariff reform, developing regulatory frameworks
and capabilities, ensuring better contract design and enforcement, and fostering transparency.
Each of these topic areas is covered in depth in the large and rapidly growing literature on
private participation in infrastructure. Surveys show that when such measures are put in place
the outcomes are better--and the public is willing to support private participation (Figure 13).

Figure 13: Opinion poll - Peru privatization of electricity generation

In favor of privatization
21%
in general

In favor of privatization
59%
if transparent process

In favor of privatization
if tariffs authorized by 65%
regulator

In favor of privatization
with investment to 69%
expand service

Source: Apoyo 2002

54. Moreover, given the large investment needs in infrastructure (Box 4), the issue is no
longer one of public or private. Meeting the financing needs will require resources from all
possible sources including through well-designed public-private partnerships and output-based
aid schemes, development of local financial markets, and better use of risk mitigation products
(Estache 2004). Opening up private participation to a wider group of players can also be
beneficial. Global investors may be withdrawing or becoming more selective because of
political and market difficulties, but this is being balanced by a growing trend towards
participation by regional and local investors. Four of the top 10 electricity sponsors in 2001-03
for example were companies from Malaysia, Brazil, Hong Kong, and Thailand (Izaguirre
2004). Participation by small-scale providers is also becoming an alternative and in some cases
a complement to conventional providers in both rural and urban areas in sectors such as water
and sewerage (McIntosh 2003).
 26

Box 4: Investment in infrastructure

Improving infrastructure networks and service delivery in developing countries requires huge amounts of
investment, estimated at 7.5 to 9 percent of GDP for low-income countries for the next five years.
Meeting these needs will require more funding than is currently allocated. For low and low middle income
countries, spending levels are up to 3 percent lower than estimated needs. The private sector accounts for
about 20 percent share of infrastructure spending, equivalent to only 10 to 15 percent of estimated annual
needs. The public sector (government, utilities) is the largest contributor, spending between 2 to 4 percent
of GDP and accounting for about 70 percent of total infrastructure spending. In recent years public
investment has declined as a result of expectations of private sector participation, fiscal adjustment
programs, and decentralization. Eight Latin American countries, for example, saw a drop from 3 percent
of GDP to 0.8 percent during the 1990s. Public investment declines have been compounded by a sharp fall
in official development assistance (ODA), which accounts for the remaining 10 percent of investment
spending. With overoptimistic expectations of private sector participation, ODA started to decline in the
1990s; infrastructure commitments from multilateral banks fell from $18 billion in 1996 to $13.5 billion
in 1999, though by 2002 they had recovered to about $16 billion.

Source: Briceño-Garmendia et al 2004; Estache 2004; Izaguirre 2004

Financial sector

55. In financial sectors, privatization is desirable but the impact on other firms and the
economy as a whole means it needs to be done carefully compared to normal commercial firms.
Experiences from countries such as Argentina, Chile and Mexico show the need for sequencing
the phasing out of state ownership over time with new private entry and improvements in the
regulatory environment. Involving reputable and experience foreign owners is also often part of
a successful strategy as they can bring the necessary skills, products, and training to the host
country and provide better allocation decisions (World Bank 2001). Minimizing government
ownership in privatized banks appears to be important. A recent study of bank privatization in
Nigeria, for example, found negative effects of minority government ownership on the
performance of many of the privatized banks (Beck et al 2005). In the banking sector more
broadly, Clarke, Cull and Shirley (2005 forthcoming) find that when government holds a
portion of the stock, especially a majority of the stock, banks continue to perform poorly.
Continued government ownership also reduces credibility, creates continued pressures for
lending toward favored borrowers, and negatively affects the earnings prospects of the banks
which then harm investors and taxpayers (Caprio 2004).

Social sectors

56. Public-private partnerships (PPPs) in social sectors such as health and education are on
the rise because of the perceived inefficiencies of the state sector and the perceived advantages
of bringing in the private sector in terms of increased competition, specialized skills, and
flexibility and innovation. Yet, there are relatively few assessments of the impacts and lessons
of PPPs in social sectors, most of which are found in health. The findings of recent reviews are
summarized in Box 5.
 27

Box 5: Public-Private Partnerships in Social Sectors

Loevinsohn and Harding (2004) review the experience with contracting approaches (mainly management
and service delivery contracts) in the delivery of primary health care or nutrition services involving non-
profit organizations. Based on ten studies of mostly low-income countries, they show that rapid
improvements were achieved in a variety of services and settings. The four cases with controlled before
and after evaluations demonstrated large impact, with the median double differences ranging from 9 to 26
percentage points. Larger double differences were observed for parameters that are easier to change, such
as immunization. Six of the ten studies compared contractor performance to government provision and all
found that contractors were more effective, both in terms of quality of care and coverage of services. The
concern that contracts are unlikely to enable service delivery on a large scale was not borne out: four of
the examples involved millions of beneficiaries. In terms of cost, non-governmental entities performed
better even when public institutions were provided similar amounts of resources. The programs were
sustainable: of the nine programs, seven have been continued and expanded. The one study which
examined equity found that contractors were able to significantly improve health services for the poor,
actually saved them money, and did a better job than government provision.

Taylor and Blair (2002) review the experience with PPPs in managing and constructing public hospitals.
In Australia, federal and state governments have introduced private participation in more than 50 public
hospitals through different approaches, including build-own-operate (BOO) transactions, conversions or
sale of hospital to a private operator, private management of a public hospital, build-own-leaseback
arrangements, and collocations in which a private wing is located within or beside a public hospital. One
example with positive results is the Mildura hospital BOO contract, awarded in 1999. Capital costs for the
new hospital were 20 percent below those for public hospitals. All performance targets have been met,
patient volumes increased by 30 percent in the first year, clinical services are provided at a lower cost, and
the operator made a profit. In Sweden, the municipality of Stockholm leased a 240-bed public hospital to a
private provider following a series of prior reforms launched in 1994. Since then, unit costs have been cut
by 30 percent and the hospital is now able to treat 100,000 more patients annually with the same
resources. Taylor and Blair show that PPPs provide innovative ways to control costs and improve service,
but they also emphasize that gains are achieved only when policymakers structure the transactions
carefully and create sound regulatory arrangements to ensure universal access, quality care, and
improvements in efficiency.

LaRocque (2004) reviews the impact of private initiatives in the education sector, an area of limited
analysis. Studies in the UK show that private finance initiatives appear to have been successful in
delivering public infrastructure in a timely and cost-effective manner. The US experience with contract
schools has been uneven, with education management organizations (EMOs) having mixed success. Early
studies showed that while they yielded benefits for students in the form of individualized instruction,
improved attendance, and greater access to computers, they had little systematic impact on academic
achievement. Recent evidence is more promising: one recent study showed that EMO-operated charter
schools registered gains in test scores over the 2000-2002 period and significantly out-gained non-EMO
charter schools with similar demographic profiles. Evidence on the impact of choice and competition on
student performance yields contradictory results: (i) some studies show that students at charter schools
perform better than public school students, while others show the reverse; and (ii) competition from
private schools improves achievement at public schools, while others show a small or no effect. As in the
health sector, experience in education shows that good contract design and a well-designed contracting
process are critical to achieving successful outcomes.

Source: Loevinsohn and Harding 2004; Taylor and Blair 2002; LaRocque 2004
 28

Bring market discipline to state enterprises

57. Policies to expose state enterprises to competition through deregulation and hard budget
constraints are an important element of any state enterprise reform effort. Allowing new private
entry forces firms to improve performance. In reforming the ports sector in Mumbai, India, for
example, the authorities opted to create a new private port rather than undertake controversial
reform involving private participation in the existing state-owned port trust JNPT. Its
operational performance improved substantially as a result of competition--and learning--from
the private sector port, which in turn is helping build constituencies for more fundamental
reforms of JNPT, including possible private sector participation down the line (Ray 2003). The
threat of failure and exit forces state enterprises to restructure and become more disciplined. It
also signals government commitment to policies of preventing continued protection to
inefficient state firms. In Bangladesh, for example, the recent closures of large loss-making jute
mills proved a powerful commitment device in this regard compared to the privatization of a
handful of smaller firms over the past several years. In Nigeria, after failed attempts to privatize
the national flag carrier, Nigerian Airways, the government decided to liquidate the company.

58. Improving the corporate governance of state enterprises is another important avenue
which has yielded success in a few cases, including electricity companies in New Zealand and
South Africa where corporate governance principles have been modeled on private sector
practices (Irwin 2004). The Organisation for Economic Co-operation and Development
(OECD) has created guidelines on the corporate governance of state enterprises, focusing on
developing an effective legal and regulatory framework, ownership policies which incorporate
transparency and accountability, and equitable treatment of and relationship with shareholders.
At the same time longstanding experience shows that such measures can be difficult to achieve
and, more importantly, sustain given the potential for continued political interference.
Moreover, effective reform of enterprises under state ownership requires that the same sensitive
issues of tariffs and employment that arise in privatization need to be addressed under state
ownership if efficiency and service delivery are to be improved.

Pay greater attention to measuring and monitoring results

59. While there are a large number of studies assessing the impact of privatization on firm
performance, and increasingly the welfare impacts, there are no systematic data on the size and
performance of the state enterprise sector as measured by share of aggregate or sectoral GDP,
share of SOE assets privatized, share of SOE losses, share of transfers and subsidies, share of
domestic credit and the like. There are also few assessments of the results of reform under state
ownership. In the absence of such data it becomes difficult to assess the extent and implications
of government ownership. Results measurement along these lines would help assess whether
objective are being achieved, increase transparency and accountability, provide a feedback loop
to correct course as needed, and play a vital role in making the case for state enterprise reform.
 29

Bibliography

Andrews, Michael, 2005. “State-Owned Banks, Stability, Privatization, and Growth:

Practical Policy Decisions in a World Without Empirical Proof.” IMF Working Paper.

Beck, Thorsten, Robert Cull, and Afeikehna Jerome. 2005. Bank Privatization and
Performance: Empirical Evidence from Nigeria. World Bank Policy Research Working
Paper No. 3511.

Briceno-Garmendia, C, Antonio Estache, and Nemat Shafik. Infrastructure Services in

Developing Countries: Access, Quality, Costs and Policy Reform. INFVP paper.

Caprio, Gerard, Jonathan Fiechter, Robert Litan and Michael Pomerleano. 2004. The
Future of State-Owned Financial Institutions. The Brookings Institution. Conference
Report No. 18.

Chong, Alberto and Florencio Lopez-de Silanes. 2003. “The Truth About Privatization in
Latin America.” Inter-American Development Bank, Working Paper No. R-486

Chong, Alberto and Florencio Lopez-de Silanes. 2004. “Privatization in Mexico.” Inter-
American Development Bank.

Clarke, George, Robert Cull, and Mary Shirley. 2003. “Empirical Studies of Bank
Privatization: An Overview.”

Estache, Antonio. 2004. Emerging Infrastructure Policy Issues in Developing Countries: A

Survey of the Recent Economic Literature. Background Paper for the October 2004 Berlin
meeting of the POVNET Infrastructure Working Group.

Estache, Antonio. 2005. “PPI partnerships versus PPI divorces in LDCs.” World Bank
Policy Research Working Paper 3470.

Garnaut, Ross, Ligang Song, Stoyan Tenev, and Yang Yao. 2005. China’s Ownership
Transformation: Process, Outcomes, Prospects. IFC Washington D.C.

Gökgür, Nilgün. 2004. Assessing Trends and Outcomes of Private Participation in

Infrastructure in sub-Saharan Africa. Draft paper prepared for PPI Conference in Africa.
 30

International Telecommunication Union. 2003.

Izaguirre, Ada Karina, 2004. “Private Infrastructure.” World Bank Viewpoint Note No.
274.

Izaguirre, Ada Karina. 2004. Private Power Projects. World Bank Public Policy for the
Pirvate Sector Note No. 281.

Kikeri, Sunita and John Nellis, 2004. An Assessment of Privatization. World Bank Research
Observer Vol. 19, No. 1.

Latin Finance.com. 2004.

Lieberman, Ira, Mario Gobbo, Anna Sukiasyan, Sumter Lee Travers and James R. Dick
Welch. 2003. “Privatization Practice Note: Europe and Central Asia Region.” World
Bank.

McKinsey Global Institute. 2001. India Country Study

McKinsey Quarterly. 2004. Corporate Governance in China.

Nellis, John, 2003. “Privatization in Latin America.” Center for Global Development.
Working Paper 31. 2003.

Organisation for Economic Co-operation and Development (OECD). 2001. “Recent

Trends in Privatization.”

Organisation for Economic Co-operation and Development (OECD). Financial Market

Trends, No. 82. 2002.

Organisation for Economic Co-operation and Development (OECD). 2004. “Privatization

in Sub-Saharan Africa: Where Do We Stand?”

Privatisation and Public Private Partnership Review 2003 – 2004. Euromoney.

 31

World Bank. 1995. Bureaucrats in Business: The Economics and Politics of Government
Ownership. Oxford University Press.

World Bank. 2000. World Development Indicators, World Bank, Washington, D.C.

World Bank. 2001. Global Development Finance 2001. World Bank, Washington, D.C.

World Bank. 2003. Trade, Investment, and Development in the Middle East and North
Africa: Engaging with the World. World Bank Washington D.C.

World Bank. 2004. Economies in Transition: An OED Evaluation of World Bank

Assistance. World Bank Washington D.C.

World Bank. 2005. Connecting sub-Saharan Africa: A World Bank Group Strategy for
Information and Communication Technology Sector Development. World
Bank/GICT Paper.