More Related Content
Similar to Political (in)stability of pension system reform (20)
Political (in)stability of pension system reform
- 1. Political (in)stability of pension system reforma Oliwia Komada (WSE and GRAPE) Krzysztof Makarski (NBP, WSE and GRAPE) Joanna Tyrowicz (IAAEU, UW, GRAPE and IZA) NBP Summer Workshop a The views presented herein are those of the authors and not necessairly those of NBP 1
- 2. Question Is pension system privatization politically stable? 1. Reform DB → partially FDC, like Poland 1999 2. Possibly Hicks-Kaldor efficient, 3. but with substantial costs today and gains only in the future. 4. Key: Is successfully introduced reform going to be stable? 5. If not one should rethink original privatization. 2
- 3. Reform DB → FDC immediate cost and delayed gains Differences in pension system deficit in % of GDP between DB and FDC scenario 3
- 4. Motivation: Reform DB → FDC, welfare effect and support The share of population gaining from introducing FDC in 1999. 4
- 5. Reform DB → FDC, after transition period everybody gains The share of population gaining from introducing FDC in 1999. 5
- 6. Motivation: BUT A number of countries reverted the privatization of the pension system: Hungary, Poland, Lithuania, Latvia, Slovakia, Czech Republic, Romania, Bulgaria (Jarrett, 2011; Schwarz et al., 2014) 6
- 7. Literature review 1. A wave of reforms Holzman and Stiglitz (2001), Bonoli and Shikinawa (2006), Gruber and Wise (2009) 2. Reform = introduce some notion of funding into the system 3. Whether or not privatization is in fact welfare enhancing Conesa and Kruger (1999), Nishiyama and Smetters (2007), Fehr (2009) 4. Despite general welfare gains... 5. ... most of these reforms got reversed Jarrett (2011); Schwarz et al. (2014) 6. Some of the reversions welfare deteriorating: Hagemejer, Makarski & Tyrowicz (2015) 7. Political economy of pension systems reforms • will the reform be implemented Cooley and Soares (1999a) • coalition between low productive workers and retirees sustain PAYGO systems Cooley and Soares (1999) • subsequent literature reviewed by de Waque (2005) or Galasso and Profeta (2002) 7
- 8. Goals and expectations Goal Suppose there already is a reform, with stable gains in the long-run: does it eventually become politically stable? Expectations • With passing of the initially old cohorts, welfare gains become majoritarian • Intend to understand/explain the reversing of reforms 8
- 9. Results - preview 1. Distribution of welfare costs and gains → abolition of funded pillar always has political support. 2. Retirees always vote for abolition. 3. No coalition low productive workers + retirees. 4. Abolition (almost) does not effect inequalities. 5. Correct assignment of property rights crucial for stability. 9
- 11. Model
- 12. Model • SOE deterministic heterogeneous agent OLG model • Households heterogenous in their preference and productivity • Competitive firms • Pension system reformed from DB to partially FDC • Taxes used to finance government expenditure and deficit in the pension system 11
- 13. Households I • Are “born” at age 20 (j = 1) and live up to 100 years (J = 80) • Belong to a type κ: • time discounting δκ • relative leisure preference 1 − φκ • productivity level ωκ • Choose labor supply l endogenously • Maximize remaining lifetime utility derived from consumption c and leisure 1 − l: Uj,κ,t = J−j s=0 δs κ πj+s,t+s πj,t cφκ j+s,κ,t+s (1 − lj+s,κ,t+s)1−φk • Subject to the budget constraint (next slide) 12
- 14. Households II • Subject to the budget constraint (1 + τc t )cj,κ,t + aj,κ,t = (1 − τl t )(1 − τ)wtωκlj,κ,t ← labor income + (1 + (1 − τκ t )rt)aj−1,κ,t−1 ← capital income + (1 − τl t )bj,κ,t ← pension income + beqj,κ,t ← bequests − Υt ← lump-sum tax 13
- 15. Financial markets and producers • Small open economy • domestic interest rate linked with global interest rate r∗ t • risk premium ξ Bt Yt (Schmitt-Grohe and Uribe (2003)) rt = r∗ t + ξ Bt Yt • Perfectly competitive firms with Cobb-Douglas production function Yt = Kα t (ztLt)1−α 14
- 16. Pension System • Pay As You Go Defined Benefit (PAYG DB) b¯J,κ,t = ρ · wt−1ωκl¯J−1,κ,t−1 • Pay As You Go Defined Contribution (PAYG DC) bI ¯J,κ,t = f I ¯J,κ,t LE¯J,t , where f I j,t = (1 + rI t )f I j−1,t−1 + τI ωκlj,κ,t • both with payroll growth indexation, rI, bj,κ,t = (1 + rI t )bj−1,κ,t−1 • Funded Defined Contribution (FDC) bII ¯J,κ,t = f II ¯J,κ,t LE¯J,t , where f II j,t = (1 + rt)f II j−1,t−1 + τII ωκlj,κ,t • with bj,κ,t = bI j,κ,t + bII j,κ,t 15
- 17. Government • Collects taxes τc, τl , τk, Υ (sum up to T). • Covers public good spending Gt = g · Nt · zt and • deficit of the pension system subsidyt = τI t wtLt − K κ=1 J j=¯Jt Nj,κ,tbI j,κ,t • Budget constraint Tt + Dt = (1 + rt)Dt−1 + Gt + subsidyt • closed with the consumption tax given by the following rule τc,t = (1 − )τfinal c + τc,t−1 + D( ˜(D/Y )t − (D/Y )final ) 16
- 18. Calibration
- 19. Macroeconomic environment Financial market and firms • global interest rate r∗ = 2% • risk premium ξ = 0.03 → domestic interest rate • deprecation d → investment rate 21% • capital share in GDP α = 0.3 • technological progress zt slows down from 3% to 1.5%, data + AWG projection Taxation • labor τl → labor income tax revenues to labor revenue • consumption τl → VAT revenues over GDP • capital τk de iure rate 19% • fiscal rule ρ i D → smooth adjustment of debt and τc • debt do GDP D/Y = 45% Pension system • contribution τ → share of benefits in GDP • replacement rate ρ → SF deficit 17
- 20. Demography Number of 20-years old Mortality rate for 60-years old ⇓ ⇓ rI ↓ rII ↓ 18
- 21. Heterogeneity - productivity ω SES Eurostat 1998, Poland Productivity ω Result: 10 values for ω 19
- 22. Heterogeneity - leisure preference φ SES Eurostat, 1998, Poland Leisure preference φ Result: 4 values for φ 20
- 23. Heterogeneity - discount factors δ • We calibrate the central value of δ to match the investment rate • We don’t have the data on stratified mortality rates or wealth • We split the model population ad hoc into 3 groups • Their discount factors are (0.98δ, δ, 1.02δ) • In total we have 120 types within each cohort • The resulting consumption Gini index in the initial steady state is 25.5, consistent with Brzezinski (2011) 21
- 24. Voting
- 25. Political economy What happens within each vote? • Policy 0 - status quo • Policy 1 - shift of contributions: funded ⇒ PAYG (Central and Eastern European countries) • Policy 2 - shift of assets • Policy 3 - a combination of the two (ex. Hungary, Poland, Bulgaria, Slovakia) 22
- 26. Pure Majority voting: the median voter • We run these votes in subsequent years independently. • If welfare gains, subcohort is in favor. • Non strategic voting: voter vote according to their preferences. • If a policy gains majority, it is put in place. • Order of voting irrelevant (we check that). Policy 1 against status quo, winner against Policy 2, winner against Policy 3 23
- 27. Results
- 28. Reform PAYG DB → FDC is Hicks-Kaldor efficient Aggregate welfare gains +1% , expressed in % of permanent consumption from the partial privatization scenario. 24
- 29. Reform DB → FDC, positive welfare effect and support The share of population gaining from introducing FDC in 1999. 25
- 30. Reform DB → FDC, after transition period everybody gains The share of population gaining from introducing FDC in 1999. 26
- 31. Nevertheless, rational voters want to withdraw it Political support for each policy change against status quo. 27
- 32. Nevertheless, rational voters always want to withdraw it Political support for each policy change against status quo. 28
- 33. Nevertheless, rational voters always want to withdraw it Political support for each policy change against status quo. 29
- 34. Shift of contribution wins, retirees prefer both changes Political support against winning scenario: contribution. 30
- 35. Vote for policy change, because of lower taxation... Differences in consumption tax relative to baseline (DB→NDC) expressed in percentage points. 31
- 36. ... due to conversion of explicit into implicit debt... Differences in debt level relative to baseline (DB→NDC) expressed in percentage points. 32
- 37. ...regardless of lower pension benefits... Differences in pension benefits relative to baseline (DB→NDC) expressed in percentage points. 33
- 38. ... which are lower due to slower accumulation in PAYG pillar 34
- 39. No coalition: low productive workers + retirees Shift of contribution, voting in 2012, consumption equivalent expressed in % of permanent consumption from the partial privatization scenario 35
- 40. Policy change slightly reduces poverty... Relative poverty defined as percentage of population with consumption below 60% of median consumption. 36
- 41. ... the effect (almost) disappears when base is kept constant Abolute poverty defined as percentage of population with consumption below 60% of median consumption from the initial steady state (stationarized). 37
- 42. The short-term improvement concerns mainly retirees Percentage of retirees with consumption below 60% of median from the initial steady state consumption. 38
- 43. Conclusions
- 44. Conclusions 1. Distribution of welfare costs and gains → abolition of funded pillar always has political support. 2. Retirees always vote for abolition. 3. No coalition low productive workers + retirees. 4. Abolition (almost) does not effect inequalities. 5. Correct assignment of property rights crucial for stability. 39