Social Protection

1
Social Protection
Public Expenditure Analysis and Manage Core
Course
Presented by Margaret Ellen Grosh HDNSP
March 21-24, 2005
Presented to
2
Social Protection

• PEAM course
• March 2005, Washington DC
• Margaret Grosh

3
What is Social Protection?

• New definition in Banks SP strategy paper
• SP as public interventions (i) to assist
  individuals, households, and communities better
  manage risk, and (ii) to provide support to the
  critically vulnerable
• Contrasts with traditional definition, as a group
  of public programs
• pensions, labor market interventions, safety
  nets, and social care

4
Boundary Issues

• Traditional SP vs other sectors SP chapters
  sticks to traditional boundaries, social risk
  management framework can be used throughout
  report and other sectoral analysis viewed through
  that lens.
• Public/private. A PER cannot cover all private
  spending in detail, but must have a notion of it
  to draw appropriate conclusions about the public
  part.
• What programs specifically? Fuzzy conceptual
  boundaries, fragmentation in institutional
  responsibility and budgets
• PER vs fuller sector work PER is selective and
  summary fiscal issues predominate, institutional
  and service delivery systems usually the least
  treated

5
Selected key issuesWhat level of spending is
appropriate?

• Significant ideological controversy
• Traditional view redistribution justified by
  moral philosophy social protection as a cost, a
  luxury
• New view as typified in social risk management
  framework social protection as an investment in
  income generation, human capital formation and
  growth

6
SP/SRM as investment

• Half or more of poverty is transient, SP can help
  reduce that substantially.
• SP helps people avoid coping strategies that
  perpetuate poverty.
• Families that can't afford a bad year can't use
  the most effective earnings strategies.
• Societies with good SP programs may be able to
  take more efficient policy choices for trade,
  industry, labor, etc.
• Societies can use good SP programs to replace
  inefficient redistributive elements in other
  programs.
• SP can help temper inequality and reduce its
  costs.
• Many safety net programs contribute to human or
  physical capital formation in addition to
  providing for current consumption.

7
Social Assistance and Social Insurance as percent
of GDP
Source Blank, Grosh, Hakim and Weigand 2004,
OECD SOCX
8
What is right assignment of resources within
sector?

• No single right answer
• Diagnostic process as defined in PRSP sourcebook
• Step 1 diagnostics on risk and vulnerability
• Step 2 look at overall balance among programs
  (including outside SP sector)
• Step 3 review individual programs performance
• Step 4 conclusions/reform plan

9
Step 1 diagnostic process

• Some very summary information from poverty and
  risk and vulnerability assessments

10
Step 2 balance among programs Example Bulgaria
11
(No Transcript)
12
Step 3 program analysis

• Will cover only selected issues in this
  presentation, more covered in guidance note

13
Efficiency example 1 unit cost analysis

• Ethiopia PER average cost per ton of food
  delivered in various safety net programs
  (excluding administration and program
  implementation costs)
• International Price 130 /mt.
• International Shipping 50
• Transport Djibouti-Regional center 65
• Local distribution transport 40
• Total cost 285
• Add in administration and even free food cost 3
  birr/kg
• Benchmark open market price of 1.5-2 birr/kg
• ? But would equivalent cash be made available?
  Is food available on these markets?

14
Efficiency example 2 inference from basic
design features

• Ethiopia PER compares public works there with
  best practice and finds shortcomings
• Value of works likely to be sub-optimal because
• Non wage costs at most 20, much lower than
  international experience for well done, diverse
  portfolio of works
• Planning process on-off separate from investment
  process
• Food typically arrives during rainy season when
  works cant be done
• Transfer gains likely to be sub-optimal because
• Cant enforce work requirement (due to rainy
  season issue) so self-targeting element weak
  (though this does reduce issue of foregone
  earnings)
• Transfer too low to affect material welfare, too
  irregular to affect risk planning
• Solutions are institutional

15
Equity analysis

• At first blush seems easy, but some real
  technical issues, to be discussed in Schwarzs
  and van de Walles complementary presentation
  within this session
• NB.
• Equity is important in all sectors
• Judgments about SP sector are based on more than
  equity.
• Methodology of equity analysis is within SP
  session because as this course is designed, each
  sectoral session includes a public good of
  methodology

16
Equity is still an issue social assistance

• Coady, Grosh, Hoddinott 2004 review 122 targeted
  transfer programs in 48 countries and find
• Moderate results on average Mean outcome
  delivers one quarter more benefits to poor than
  would universal transfer
• Very much better results in best programs top
  ten deliver two to four times more benefits to
  poor than would universal transfer
• Significant targeting failures one quarter of
  targeted programs are regressive.

17
Equity is still an issue pensions
Source De Ferranti et al. 2004, Figure 9.9
18
Equity is still an issue pensions

• Mexico results not unusual
• But is the comparison fair?
• If payments are deferred compensation (earnings)
  then poverty targeted expenditures the wrong
  benchmark.
• Still issues of equity across generations,
  genders, income levels, work histories

19
Pension System Equity in MENA
Comparison of Benefit Rates Across Different
Income Groups

• In countries like Iran, Egypt, and Yemen, low
  income workers earn more in retirement than while
  working
• Higher income workers in Egypt earn much lower
  pensions relative to their income level than low
  income workers even though their contributions
  were much higher
• Few incentives to contribute in these countries
  evasion is high

20
Pension System Equity in MENA
Comparison of Benefit Rates Across Different
Income Groups

• In countries like Iran, Egypt, and Yemen, low
  income workers earn more in retirement than while
  working
• Higher income workers in Egypt earn much lower
  pensions relative to their income level than low
  income workers even though their contributions
  were much higher
• Few incentives to contribute in these countries
  evasion is high

21
Managing risk a tension

• PERs very concerned with fiscal risk
• Adequate SRM and SSN implies programs with
  entitlement access
• Argentinas Trabajar program vs Maharasthras
  Employment Guarantee Scheme
• very rare in practice because of fiscal issue
  (and sometimes administrative constraints)
• Even counter-cyclicity rare in LAC for each 1
  loss of GDP, the amount of targeted spending per
  poor person declined by 2 (de Ferranti, et al
  2000)

22
Fiscal Sustainability in Pensions
Brazil Critical Social Security Issues, June
2000.
23
Summary of SP in PER Guidance Note

• Sector wide view
• Very brief synopsis of poverty, risk and
  vulnerability
• Overview of budget allocation, trends, processes
• Individual program analysis
• Adequacy
• Equity
• Efficiency
• Contribution to risk management
• Delivery mechanisms
• Sustainability
• Impact

24
Hallmarks of good analysis

• Numbers clearly defined and sources given.
• Uses benchmarks extensively (not just on
  expenditures but on inputs, prices, outputs,
  ratios among these)
• Chooses benchmarks wisely (e.g. neighboring
  countries, countries of similar income, others
  the country wants to emulate or adjusts for
  differences in demographics or poverty profile)
• Contrasts trends and point in time as applicable
• Conveys enough of the storyline and details to
  persuade reader of recommendations
• Crafts together story from available sources and
  literature outside of PER.