ASSESSING THE SOCIAL IMPACT OF PRIVATIZATION John Nellis Center for Global Development Washington, D

1
ASSESSING THE SOCIAL IMPACT OF PRIVATIZATION
John NellisCenter for Global
DevelopmentWashington, DC
2
MOST STUDIES RATE PRIVATIZATION A MICROECONOMIC
SUCCESS
3
PROFITABILITY, EFFICIENCY RETURNS TO
SHAREHOLDERS GENERALLY INCREASE
4
MACROECONOMIC IMPACT POSITIVE (IMF)

• Net proceeds 2 GDP generally saved, not
  spent
• Growth impact positive (?)
• Good proxy for liberalizing reform
• Financial flows to govt. often increase
  post-privatization

5
Continuing debate on the extent ownership
change--- or other factors--- explains
performance improvements
6
PRIVATIZATION HIGHLY INCREASINGLY
UNPOPULAR---- IN LATIN AMERICA, SOUTH ASIA,
AFRICA TRANSITION COUNTRIES
7
(No Transcript)
8
Sri LankaAttitudes Towards Privatization (2000)
9
IN RUSSIA, 2/3 INTERVIEWED LOST MORE THAN
GAINED FROM PRIVATIZATION 2001 1600
respondents only 5 said opposite
10
PRINCIPAL SOCIAL CRITICISMS OF PRIVATIZATION

• UNFAIR IN CONCEPTION DESIGN
• BENEFITS RICH, FOREIGN CORRUPT
• INCREASES INEQUALITY POVERTY

11
QUESTIONS

• IS PRIVATIZATION INCREASING INEQUALITY?
• IF SO, HOW TO WHAT EXTENT?
• WHAT CAN SHOULD BE DONE ABOUT IT?

12
ISSUE UNDER REVIEW BY

• UNU/WIDER
• IADB
• CGD
• WORLD BANK
• INDEPENDENT SCHOLARS

13
HOW MIGHT PRIVATIZATION AFFECT EQUALITY?

• DISTRIBUTION OF ASSETS
• EMPLOYMENT RETURNS TO LABOR
• ACCESS (COVERAGE) PRICES
• FISCAL POSITION RESOURCE ALLOCATION OF
  GOVERNMENT

14
MOST STUDIES CONCENTRATE ON EMPLOYMENT
ACCESS/PRICES
15
MOST STUDIES FROM LAC

• LARGE AMOUNT OF INFRASTRUCTURE PRIVATIZATION
• HOUSEHOLD EXP. CONSUMP. SURVEYS AVAILABLE
• LARGE OF LOCAL RESEARCHERS

16
FEW ATTEMPT FULL COUNTERFACTUAL

• DATA LIMITATIONS
• FEW SIMPLE ASSUMPTIONS (e.g., no price
  changes)
• RELIANCE ON BREAK POINTS IN PREVIOUS TRENDS
• HINT THAT ELABORATE COUNTERFACTUALS SOMEWHAT
  SUBJECTIVE

17
M. Torero A. Pasco-Font, Social Impact of
Privatization Regulation of Utilities in Peru.
WIDER DP 2001/17D. Mckenzie D. Mookherjee,
Distributive Impact of Privatization in Latin
America Evidence from Four Countries, draft,
BU, 2003J. A. Delfino A. A. Casarin, Reform
of the Utilities Sector in Argentina, WIDER DP
2001/74G. Barja M. Urquiola, Capitalization
and Privatization in Bolivia An Approximation to
an Evaluation, IADB-CGD paper, 2003
18
MCKENZIE MOOKHERJEE TRY TO

• MEASURE IMPACT ON ACCESS, PRICE QUALITY
• CALCULATE VALUE OF CHANGES FOR CONSUMERS ACROSS
  INCOME DECILES
• MEASURE CONSEQUENCES FOR INEQUALITY POVERTY

19
Universe 10 infrastructure privatizations in
Argentina, Bolivia, Mexico, Nicaragua
20
Use of surveys poses problems

• Report expenditure, not price info
• Most limited to urban households
• Limited number (2 4)

21
Forces simplifying assumptions, e.g.

• Use aggregate price indices
• Estimate demand elasticities
• Assume rural responses match urban
• Assume few observations yield trend
• Assume laid off workers never re-employed (upper
  bound)

22
FINDINGS

• ACCESS UP IN ALL CASES
• WATER, ELECTRICITY EXPANSION PARTICULARLY
  BENEFICIAL TO POOR
• T-COMM EXPANSION TO MID-TOP OF DISTRIBUTION
• PRICES UP IN 5 CASES, DOWN IN OTHER 5
• SERVICE QUALITY IMPROVES MARKEDLY

23
QUALITY SHIFT CAN BE VERY IMPORTANT

• Argentina infant mortality down 5 to 7 in
  areas where water privatized
• Poorer the area, greater the decline (up to 24)
• (Galiani, Gertler, Schargrodsky, 2002)

24
WELFARE CHANGES

• Infrastructure costs small part of normal
  household budget---effects small
• Value of access outweighs price increases
• Water price rises neg. affect welfare---again,
  effects small

25
IMPROVED ACCESS

• Peru telecom 167
• electricity 33
• Bolivia telecom 123
• electricity 2.7
• water 15
• Argentina telecom 30
• electricity 11
• natural gas 30

26

27
(No Transcript)
28
EVEN WHEN WELFARE , POOR CONSUMERS CAN DO
SUFFERE.G., ENDING ILLEGAL HOOK-UPS
(Argentina)
29
Ownership effects unlikely to effect bottom end
of income distributionEmployment/consumer
effects more important
30
EMPLOYMENT

• EMPLOYEE s DECLINE BEFORE AFTER SALE
• 50 loss rate in Argentina Mexico

31
SURVEY OF 308 PRIVATIZED FIRMSEMPLOYMENT LOSS
IN 79EMPLOYMENT NEUTRAL OR GAIN IN 21(Chong
Lopez-de-Silanes, 2002)
32
EMPLOYMENT

• RETAINED EARN ABOUT SAME
• WORK MORE HOURS LESS SECURITY
• MEN, YOUTH, BETTER EDUCATED THE WINNERS WOMEN,
  THOSE gt 45 THE LOSERS

33
BUT.. DISMISSED SMALL OF WORKFORCE
PRIVATIZATION NOT PRIME CAUSE OF HIGH
POST-REFORM UNEMPLOYMENT
34
FISCAL EFFECTS

• Positive flow of funds (despite
  underpricing)
• More from end of subsidies new corporate taxes
  than from sales proceeds
• Public debt down social expenditures up in
  many cases
• Privatization a fiscal opportunity

35
CONCLUSIONIN SHORT RUN, PRIVATIZATION WORSENS
DISTRIBUTION HEIGHTENS PERCEPTION OF
UNFAIRNESS
36
SO WHAT?
37
RISING INEQUALITY THE NECESSARY (HOPEFULLY
TEMPORARY) PRICE TO PAY FOR PUTTING THE ECONOMY
BACK TO WORK?
38
Wealth effects mainly important in
transition economies Income effects small
perhaps temporary Increased access
outweighs price increases
39
Poor sometimes primary beneficiaries More
often, all benefit but upper deciles more
than lower General welfare increases,
inequality as well
40
In best-studied Latin American cases.

• privatization has a very small effect on
  inequality
• changes to Ginis 0.02 or less
• Privatization either reduces poverty or has no
  effect on it..

41
SEVERELY NEGATIVE PUBLIC PERCEPTION NOT
SUPPORTED BY EMPIRICAL ANALYSES