Reforming pensions in Chile Nicholas Barr and Peter

Reforming pensions in Chile Nicholas Barr and Peter Diamond London School of Economics Massachusetts Institute of Technology Pension reform in Chile, American Economic Association Annual Meetings Chicago, 6 January 2017

Reforming pensions in Chile 1 Backdrop 2 The 2008 reforms 3 The 2015 Pension Commission (the Bravo Commission) 4 The Global Proposals Nicholas Barr and Peter Diamond, January 2017 1

1 Backdrop Nicholas Barr and Peter Diamond, January 2017 2

Thumbnail history • Chile introduced mandatory funded individual accounts in 1981, provided by specialist firms, Administradoras de Fondos de Pensiones (AFPs) • Widely followed in Latin America and, from the late 1990 s, Central and Eastern Europe and China • Retreat from individual accounts in many countries Nicholas Barr and Peter Diamond, January 2017 3

Objectives of pension systems • The primary objective of pensions is economic security in old age • Achieving that objective requires • Poverty relief • Consumption smoothing across a person’s lifetime • Insurance against low income and wealth in old age • Cannot wholly achieve all the multiple objectives simultaneously; thus policy has to optimise across all, not maximise any particular one Nicholas Barr and Peter Diamond, January 2017 4

Strategic critique of 1981 Chile design • Individual funded accounts are complex to implement • If well implemented they provide consumption smoothing but little risk sharing or poverty relief • Thus 1981 arrangements were not a pension system but only part of a pension system Nicholas Barr and Peter Diamond, January 2017 5

2 The 2008 reforms • Two strategic reforms introduced as a result of the Marcel Commission (2006) Nicholas Barr and Peter Diamond, January 2017 6

Non-contributory solidarity pension • Based on test of age and residence • Targeted on bottom 60% of the pensioner income distribution • Covers urban, rural, formal and informal, employed and self-employed • Turned the system into a genuine strategy by including poverty relief Nicholas Barr and Peter Diamond, January 2017 7

Auction mechanism • To tackle problem of continuing high administrative charges • Accumulation • AFPs bid for all new entrants over a two-year period • Have to stay with winning bidder for at least 18 months • AFP has to offer the same low commission to its existing members and to any new members • Benefits: • Endorsed the Online Pension Consultation and Bidding System (SCOMP) introduced in 2004 • Mandatory to consult SCOMP, which includes insurance companies bidding for individual retirees Nicholas Barr and Peter Diamond, January 2017 8

3 The 2015 Pension Commission (the Bravo Commission) • The 2008 strategy was sound because it addressed the three core objectives of poverty relief, old-age income insurance, and consumption smoothing • But it left unfinished business which the Bravo Commission was set up to address Nicholas Barr and Peter Diamond, January 2017 9

3. 1 Bravo Commission: diagnosis of problems • Known problems of funded individual accounts • Low coverage, largely because of informality • Inadequate pensions, largely because of incomplete contributions records • Continued high charges • Exposure of plan members to risk • Gender inequality: quadruple whammy • Hostility towards the AFP system (though unclear whether the root cause was pension design or the level of benefits) Nicholas Barr and Peter Diamond, January 2017 10

3. 2 Bravo Commission: specific proposals 58 specific proposals, only outlined here • Expanding the solidarity pension system • Increasing contributions, including (for the first time) an employer contribution • Increasing competition within the AFP system • Raising earliest pension age • Improving gender equity • Changing arrangements for drawdown • Removing exemptions, e. g. for the military Nicholas Barr and Peter Diamond, January 2017 11

3. 3 Bravo Commission: Global Proposals • Global Proposal A: retain current strategy and work within it to improve it • Global Proposal B: partial move to PAYG • Global Proposal C: full move to PAYG Nicholas Barr and Peter Diamond, January 2017 12

4 The Global Proposals Nicholas Barr and Peter Diamond, January 2017 13

4. 1 Global Proposal A (12 votes) Builds on the 2008 reform by • Expanding the non-contributory system by • Increasing the solidarity pension • Financing the increase through increased taxes and a new 2% employer contribution to the solidarity fund • Reforming the savings element by introducing • A new 2% employer contribution to AFP accounts • A new government AFP to strengthen competition • Further action to reduce charges, e. g. extending auction mechanism • Improving gender equity by • Over time, equalizing pension ages of men and women • Mandating the use of unisex life tables • Motivation: to increase benefits immediately while preserving sustainability Nicholas Barr and Peter Diamond, January 2017 14

4. 2 Global Proposal B (11 votes) • Diverts contributions on earnings up to 350, 000 pesos (USD 520)/month (about half of total contributions to AFPs) into a new partially-funded element, ‘organized through citizen social security accounts (which may take the form of notional accounts)’ • Redesigns the solidarity pillar • Higher non-contributory pension for bottom 80% • Tax-financed match to the new social security accounts for people with low pensions • Financing the match is ‘a three to four percent increase in the contributions from employers, and with complementary support from state contributions’ • Motivations • To address widespread hostility towards AFP system • To increase (and help finance) pensions immediately Nicholas Barr and Peter Diamond, January 2017 15

4. 3 Global Proposal C (1 vote) • Absorbs both the stock of assets of the AFP system and the flow of contributions into a new DB system • ‘Global Proposal C seeks to respond to the views expressed during the public participation process by transforming the system into one that is purely pay-asyou-go’ • ‘The full amount of contributions will be used as income to pay contributory pensions, with no State subsidies…. On average, retirement benefits would be increased by 75 -100%. Contribution rates will not be raised until 2035’ • Motivations: • Higher benefits • An ideological objection to the AFP system Nicholas Barr and Peter Diamond, January 2017 16

4. 4 Assessment • Proposal A is the best way forward • Proposals B and C are mistaken Nicholas Barr and Peter Diamond, January 2017 17

Why Proposal A • The 2008 strategy is sound • The various parts of the system fit together to achieve the multiple objectives of poverty relief, insurance and consumption smoothing • The solidarity pension provides poverty relief and some insurance • Individual accounts mainly provide consumption smoothing • Long, stable history • Thus very different from, e. g. , Argentina, China, Hungary and Poland Nicholas Barr and Peter Diamond, January 2017 18

Why Proposal A (2) • Good base for expanding the solidarity pillar • Range of flexible options for increasing pensions – Gains formal and informal workers, including women – Option of greatest gains for those with smallest AFP pension • Thus higher pensions within chosen fiscal envelope • Can achieve distributional effects similar to Proposal B • Individual accounts also a sound base for reform • Adequate replacement rate for workers with fairly complete contributions records • Government AFP with default option • Option of central clearing house for record keeping, transferring contributions to funds and policing against evasion (as in Sweden) • Option of move towards a lower cost plan, e. g. the default in the Swedish Premium Pension or the US Thrift Savings Plan • Maintains saving – important for reasons below Nicholas Barr and Peter Diamond, January 2017 19

Problems with Proposal B • Focus mainly on the formal sector • NDC + positive match benefits formal-sector workers • Highest benefits for workers with high contribution densities • Low AFP pensions mainly for people with incomplete contribution records, a particular problem for women; NDC does not address this problem • Lack of specificity, hence • Hard to judge distributional effects, though with strong suspicion of raiding the kitty, i. e. concealed redistribution from future workers and pensioners • Hard to judge finance (presentation is in terms of the system when mature, with no discussion of the transition) • Vulnerability to political pressure to raise $350 K ceiling • Government might not resist such pressures since a further move towards PAYG can create short-term fiscal windfall • Hence risk that Proposal B will approach Proposal C • Insufficient attention to significant additional administrative costs (0. 41% in NDC system in Sweden, very much a lower bound) Nicholas Barr and Peter Diamond, January 2017 20

Problems with Proposal B: Effect on saving • Transferring benefits from future retirees to current ones risks a negative impact on aggregate saving • Declining fertility will lead to a smaller workforce • Rational response is to make each worker more productive through increased investment in human and physical capital • Reforms that reduce saving affect investment and output growth, hence fundamentally the wrong way to go in Chile • For such reasons, three Commissioners (Attanasio, Meghir and Mitchell 2015) wrote: ‘Our objection to Proposal B, which in the short run would provide a level of pensions very similar to that of Proposal A, was about how … pensions would be financed. Proposal B would divert money from workers’ individual accounts and give the cash to current pensioners, rather than building up funding for their own retirement. […] [B]y diverting savings, this plan would undercut investment and future productivity growth. ’ Nicholas Barr and Peter Diamond, January 2017 21

Problems with Proposal C • Rejected by Commission because • Transfers worker’s accumulations to a PAYG system without compensation • Is unsustainable: ‘The proposal requires an abrupt and large increase in … contributions and taxes [once] … the reserves are depleted’ • Puts saving and investment at risk to a greater extent than B Nicholas Barr and Peter Diamond, January 2017 22

Consistency with past positions • Peter Diamond and I have criticised World Bank’s over-promotion of Chile model for over 25 years • Thus our support for Proposal A is not ideological • We support Proposal A because of • Its flexibility over the distribution of benefits • Its virtue in preserving investment levels • We take account of several sets of facts • The system has been in place for 35 years, has improved its design over time and has mostly met the transition costs • Fertility is declining • There are options within the existing strategy for introducing simple and cheaper ways of saving Nicholas Barr and Peter Diamond, January 2017 23

Consistency with past positions (2) • This line of argument does not imply that we regard individual funded accounts elsewhere as necessarily worthwhile – the capacity to implement and/or finance the approach is absent in many countries • For such reasons we opposed individual funded accounts in our 2005 and 2010 reports for the Chinese government (Barr and Diamond 2010 b). • That we take a different position under different circumstances is not inconsistent – if circumstances differ, so does the appropriate policy Nicholas Barr and Peter Diamond, January 2017 24

References Barr, Nicholas (2012), The Economics of the Welfare State, OUP, Ch. 7 Barr, Nicholas and Diamond, Peter (2009), ‘Reforming pensions: Principles, analytical errors and policy directions’, International Social Security Review, Vol. 62, No. 2, 2009, pp. 5 -29 (also in French, German and Spanish) Barr, Nicholas and Diamond, Peter (2010 a), Pension reform: A Short Guide, New York and Oxford: OUP, also in Spanish, Chinese and Polish Barr, Nicholas and Diamond, Peter (2010 b), Pension Reform in China: Issues, Options and Recommendations, China Economic Research and Advisory Programme, February 2010, http: //econ. lse. ac. uk/staff/nb/Barr_Diamond_China_Pensions_2010. pdf Barr, Nicholas and Diamond, Peter (2016), ‘Reforming pensions in Chile’, Polityka Społeczna, No. 1, 2016, pp. 4 -9, http: //econ. lse. ac. uk/staff/nb/Barr_and_Diamond_2016_Chile. pdf Executive Summary: Presidential Advisory Commission on the Pension System 2015 (the Bravo Commission) (2015), www. comision-pensiones. cl/Documentos/Get. Resumen. Ingles Nicholas Barr and Peter Diamond, January 2017 25