ASEZA Strategic Plan

1
Pension Reform in Eastern Europe and
Eurasia Experiences and Lessons Learned
David Snelbecker The Services Group USAID
Workshop for Practitioners on Tax and Pension
Reform Washington, DC, 27-29 June 2005
2
Pension system objectives

• To what extent do reforms in Eastern Europe and
  Eurasia achieve social, fiscal, and financial
  objectives?
• Lifetime consumption smoothingset aside money in
  work years to receive back in retirement years.
• Insuranceuncertainty of longevity, disability,
  survivorship systemic risks (markets and
  institutions in funded systems demographic,
  political, and fiscal risks in unfunded systems)
• Redistribution and social safety net
  provisionincluding support for the poor
• Compatibility with economic growthincluding
  labor market efficiency, capital market
  development, and overall efficient social
  programs
• in particular, Hungary, Poland, Kazakhstan,
  Kosovo, Ukraine

3
GDP per capita
(for 2003, in constant 2000 USD)
Source World Development Indicators, World Bank
4
Average growth rates
Source World Development Indicators, World Bank
5
Financial sector depth
Source World Development Indicators, World
Bank, 2003
6
Demographics
Source CIA World Factbook
7
Pension reform in Hungary

• Scaled back PAYG componentaccrual rate
  1.65?1.22 of earnings contribution rate 30?22
  of wages retirement age 55w/60m?62 Swiss
  indexing total social contributions41 of labor
  costs, plus PIT
• Introduced funded component6?8 of wages
  contributed 18 mandatory mutual benefit funds
  managed by members (81 voluntary funds) outside
  asset managers voluntary participation in the
  new system supervised by Financial Supervisory
  Authority
• Issues erosion of PAYG reforms can reduce
  long-term fiscal sustainability labor burden
  still high high admin fees investment issues

8
Pension reform in Poland

• Introduced Notional Defined Contribution first
  pillar (PAYG) notional interest rate of 75?100
  of real wage base growth contribution of 12.22
  of wages to NDC total social contributions32
  of labor costs, PIT
• Introduced funded component15 open pension
  funds managed by universal pension societies
  contribution rate7.3 of wages centralized
  collection through ZUS
• Issues admin/IT issues have impeded NDC account
  statements labor burden remains high high admin
  fees investment issues

9
Pension reform in Kazakhstan

• Phasing out PAYG componentnew unified social
  contribution20 of gross wage, plus PIT minimum
  pension guarantee for old system
• Introduced funded component15 private funds plus
  State Accumulation Fund contribution rate10 of
  wage
• Issues how to increase coverage, support the
  poor, and achieve adequate replacement rates for
  those not working full career in long run
  investment issues

10
Pension reform in Kosovo

• Replaced old PAYG low-coverage system with
  universal basic pensionfunded from general
  revenues paid through banking system universal
  coverage
• Introduced funded componentcontribution rate10
  of wages single provider regulated by central
  bank
• Issues maintaining sound governance through
  political status discussions addressing issues
  of domestic investment

11
Pension reform in Ukraine

• Slightly scaling back PAYG pillarcontribution
  rate 33?26 of gross wage total social
  contributions29 of total labor cost, plus PIT
• Will introduce funded component7 of gross
  wages centrally managed by existing Pension
  Fund, contracting out to asset managers
• Issues ensuring fiscal sustainability in short
  run in order to introduce funded component, and
  in long run administrative challenges in
  introducing funded system investment issues

12
USAID assistance throughout the region

• Bosnia
• Bulgaria
• Croatia
• Hungary
• Kazakhstan
• Kosovo
• Lithuania
• Macedonia

• Montenegro
• Poland
• Romania
• Russia
• Serbia
• Slovakia
• Ukraine
• OECD-INPRS

13
Low initial real net returns in FF
ILO overview Poland Kosovo
Sources Fultz (2004) Polish Ministry of Social
Policy KPST authors calculations
14
Real returns in PAYG components
Poland NDC (actual) PAYG hypothet-ical returns
Sources Polish Ministry of Social Policy CIA
World Factbook
15
Labor market distortions
High social contributions (top right), of which a
small share is now for funded pensions (bottom
right), plus personal income taxes (bottom left),
adds up still to a high burden on labor.
Sources World Bank James (2005) KPMG
Investing in guides (2005)
16
Unemployment and shadow economy
Unemployment
Shadow economy
Sources World Development Indicators Schneider
(2004).
17
Risk management
18
Inter-generational redistribution

• Pre-reform winners and losers Current
  generations receiving unsustainable benefit
  levels given contribution rates are winners.
  Future generations that face either benefit
  curtailments or contribution hikes are losers.
• Winners and losers from reform Elderly and those
  near retirement should be unaffected. Recipients
  of individual accounts win. Payers of transition
  costs (higher taxes or lower benefits, today or,
  if debt is issued, tomorrow) lose. Payers of
  lower payroll taxes win.
• Society wins overall (Pareto improvements) only
  if there are efficiency gains. Otherwise reforms
  simply re-allocate between generations.
• Generational accounting can help model
  inter-generational effects.

19
Intra-generational redistribution

• Support to the poorminimum pension guarantees
  (in PAYG or FF components, or both) vs. universal
  pensions vs. non-pension programs
• Expensive pension programs can crowd out social
  programs for the poor
• Increased coverage better reaches the poor
• Replacement rates indicate pension adequacy
• Disability and survivor benefits mostly
  unreformed
• Gender issues are complex

20
Do pensions crowd out other social spending?
Share of elderly in the poor (lowest 33), and
in total population Expenditures on
pensions and assistance for poor
Sources Aguirre International OECD social
expenditure database Kazakh, Ukrainian, Kosovo
finance ministries
21
Replacement rates, pre-reform
Average pension / average wage
22
Gender issues

• Increases and unification in retirement ages can
  affect women most since women retire roughly five
  years earlier pre-reform.
• Since women disproportionately make up oldest
  generations, retirement age increases are to
  large extent transfers from middle-aged women to
  elderly women.
• Survivor benefits are particularly important for
  women.
• Since women tend to work in formal labor fewer
  years than men, they might not get adequate
  pensions from funded systems.
• Joint annuities can be required in law, also
  requiring unisex mortality tables.

23
Macroeconomic issues

• Savings and investmentlinks to growth
• Efficiency gainsmostly in labor markets
• Financial marketsa chicken-and-egg problem
• Balance-of-paymentsoverseas investments can have
  positive impact in very long-term but mixed in
  the near term
• Dynamic general-equilibrium overlapping
  generations (OLG) models can help assess macro
  effects

24
Pension expenditures
High pension expenditures pre-reform ( of GDP,
top left), are being stabilized or slightly
falling in some reforms (Poland, below left),
and significantly falling in other reforms
(Kazakhstan, Below right).
Sources IMF Chlon-Dominczak (2004) Andrews
(2001)
25
Outstanding implicit pension liability, pre-reform
Sources World Bank, assuming 5 discount rate
26
Transition costs

• Current taxes can be raised, on labor or
  elsewhere
• Pension or other expenditures can be cut
• Debt can be issued, to be repaid by future tax
  increases or expenditure cuts (beware of too much
  high-interest explicit debt replacing
  low-interest implicit debt)
• Efficiency gains can be sought

27
Projected pension assets, 2020, of GDP
Sources Holzman and Hinz (2005)
28
Country riskinessSPs credit ratings
29
Laws, regulations, rules, institutions for sound
funded tiers

• Investment guidelines directed toward liquid,
  publicly traded and regulated assets (generally
  stocks and bonds)
• Investment for the sole and exclusive purpose of
  preserving assets and increasing return in the
  interest of participants
• Governance procedures that insulate investment
  decisions from political issues
• Diversification, including sector, risk profile
  and country
• Separation of asset custody from asset management
  
• Legal environment where individual rights can be
  asserted and protected, including shareholder
  rights and pension participant rights
• Institutional capacity to accomplish investment
  and report investment results
• Transparency and supervision

30
Administrative fees
Sources Anusic (2004)
31
Reduction in assets and returns
Sources Anusic (2004)
32
An average Polish worker in new system (PLN)
Sources Polish Ministry of Social Policy (2004)
33
Conclusions and lessons learned

• Reforms have meaningfully improved long-term
  fiscal sustainability and inter-generational
  equity. Nonetheless, further fiscal reforms in
  many cases still are needed.
• Coordinated attention is needed to reduce the
  total labor burdencontinued pension reform,
  reform of other social programs funded from wage
  taxes, and personal income tax reform.
• Administrative costs have been significant and
  should be considered at the reform design
  stagefunded pillars should be designed large
  enough to achieve economies of scale, with
  cost-efficient administration, or alternative
  reform designs should be chosen.
• Introduction of funded components or NDC reduces
  redistribution, meaning policymakers need to
  consider how best to help the poor and those who
  do not work full careers through complementary
  programsuniversal benefits vs. needs-tested
  survivor and disability unisex actuarial table
  for annuities non-pension social programs.
• Supervisory regimes of funded pensions need
  ongoing monitoring and revising, addressing
  issues that arise as pension systems and
  financial markets developinvestment restrictions
  (including abroad) excess liquidity herding
  etc.

34
Conclusions and lessons learned

• Risks can be managed in different
  waysdiversification across financial instruments
  within a DC mandatory pillar across funded and
  PAYG components or across voluntary savings.
  Deciding which approach best maximizes
  risk-adjusted returns requires quantitative
  analysis.
• Institutions and implementation are as important
  as policy design in ultimately determining reform
  success.
• How transition costs are paidtax increases,
  expenditure cuts, debt, or efficiency gainsis at
  least as important in overall reform impact as
  the final-state design.
• Assistance must build capacity among policymakers
  and civil societya lengthy, inclusive
  policymaking process is needed, after which its
  time to close the sale.
• The most important lesson of sequencing is to
  seize a political window of opportunity when it
  opens, designing reforms in the context of level
  of country development.