Matching Contributions for Pensions: A Review of International Experience. Prof. Robert Holzmann University of Malaya, CEPAR, CESifo, IZA

Transcription

1 Matching Contributions for Pensions: A Review of International Experience Seminar & Book Launch Lecture Hall 3, Faculty of Business and Accountancy (FBA) University of Malaya, April 11, 2013 Prof. Robert Holzmann University of Malaya, CEPAR, CESifo, IZA

2 Background Coverage of the workers/elderly under a mandated and/or voluntary pension scheme is quite likely the ultimate challenge of social policy Disappointment with coverage expansion under reformed earnings-related mandated schemes Concerns with coverage expansions through (generous and tax-financed) social pensions Limited take-up of voluntary funded provisions for old-age even if tax sponsored

3 and Motivation Hence central idea: Subsidizing participating in a contributory scheme to increase coverage and/or benefit adequacy Is getting increasing traction in countries across development levels: HICs to LICs But little consolidated knowledge and if experience from few high-income countries A first stocktaking and policy conclusions through Workshop on DC in June 2011 Teaming-up with BBVA and RIPPA Revised&additional contributions as book Complements coverage book 2009

4

5 Road Map I. Objectives, Definitions, Mechanisms, Modalities II. III. IV. MDC Approach under Main Reform Scenarios: Examples and Questions Country Lessons across Income Levels Policy Conclusions and Next Steps

6 Key Objectives I. Objectives, Definitions, Mechanisms, Modalities Coverage: Increase in basic or supplementary benefit coverage of population against key social risks (in particular old age and health) Informality: Reduce incentives for informal sector participation (i.e. evading costs of participating in mandated social insurance schemes) Fiscal efficiency: under the same coverage/adequacy target to be fiscally more efficient compared to o universal and means-tested (ex-post) transfers or o transfers to cover deficits in earnings-related schemes Other possible secondary objectives, e.g. o o Redistribution: Correcting unintended redistribution of other financing methods, better alignment of incidence of costs through individual contributions Prefunding: Facilitating transition to pre-funding of benefits by inducing individual contributions and establishing feasible funding stream 6

7 M(D)C Core Elements (Definition) Contribution based scheme, with contribution by the beneficiary expected, at least in the long run Direct financial contributions by employer and/or government to encourage individuals scheme participation (in contrast to indirect expenditure/tax incentives) Secondary design features include major elements of nudging (such as auto-enrolment and default contribution levels), financial education, social marketing, advocacy et al Schemes can be funded or unfunded (typically fully funded and DC type for pensions and (mostly) unfunded and basic package type for health) Schemes can be mandated or voluntary, with mandated schemes typically focused on extension of coverage, and voluntary schemes can focus on coverage extension (if coverage low) or adequacy (if coverage given but generosity low) 7

8 Envisaged Mechanism (Theory of Change) External conditional contribution to overcome nonparticipation decision due to, e.g. o High discount rate o High cost to belong to the formal economy o Liquidity constraints, etc Other interventions to overcome behavioral limitations through, e.g. o Nudging and other interventions to improve decision architecture (e.g. default options) o Social marketing, edutainment and other advocacy interventions to change behavior in non-cognitive manner Nonparticipation decisions in mandated scheme due to high tax element, high administrative costs, lacking credibility, etc should be directly addressed! 8

9 Design Modalities (example pensions) Pure matching contribution rate, say 20, 50 or 100 percent of own contribution, with possible multi-layer or tiered structure; or flat rate matching contribution in local currency with possible multi-layer or tiered structure Threshold (ceilings/caps) on contribution base, or overall matching contribution amount Eligibility conditions for matching, such as income level, family status, number of and age of children, employment status, SME, non regular workers (agriculture sector, fishing), etc. Matching contribution (ex-ante transfer) plus minimum pension guarantee (ex-post transfer) conditional on prior contributions (length or level) Conditional withdrawals for housing purchase, liquidity constraints, unemployment, etc. Consumption tax treatment benefits only subject to income tax at time of receipt similar to other pension specific savings Payout defaults (eg minimum annuity levels, deferred annuities) 9

10 Expected and likely outcomes (introduction or expansions through higher matching/threshold) Politically expected Coverage increase - participation of prior non-participants Increase in saving and benefit levels of participants within the scheme Non-decrease of savings in other programs Leveraging of fiscal costs - Net increase in national savings exceeding fiscal costs of match 10

11 Expected and likely outcomes (2/2) Economically expected (derived from first principles) Participation increase mostly positive or zero (due to substitution effect) Scheme savings effects mostly ambiguous (due to substitution and income effect) but at times clearly negative or positive (e.g. threshold increase and individuals below/at old, and above old/new threshold) Overall (scheme and non-scheme) savings effects fully ambiguous (theoretically and empirically) 11

12 Administrative Modalities Contribution collection by employer, financial market institution, social security institutions, and local aggregators (NGOs, etc). Record keeping and client communication by financial market institution, regional or national social security funds Asset management by pension/health funds; specialized asset manager; regional or national social security institutions. Benefit disbursement by financial market institution; regional or national social security social security institutions 12

13 II. MDC Approach under Main Reform Scenarios: Examples and Questions 1. Supplementing universal (basic or meanstested) benefits 2. Supplementing (reduced) earnings-related benefits 3. Coverage expansion within the mandated SI scheme 4. Coverage expansion outside mandated SI schemes universal approach 5. Coverage expansion outside mandated SI schemes sector/group specific approach 13

14 1. Supplementing universal benefits (basic or means-tested) Done in few countries such as New Zealand (demogrant): KiwiSaver with auto enrollment and constrained opt out UK (Beverige flat contribution/benefit scheme): Saving Gateway for low income earners (plans canceled), now NEST [Australia, Denmark, Netherlands, Mauritius, South Africa] Could become more important as earnings related schemes are reduced or, perhaps, eliminated: As supplement to basic provisions As means to address the transition problem As new integrated contributory approach to assure adequacy Key questions are effectiveness in supplementary coverage and fiscal terms compared to alternatives Mandated/voluntary schemes 14

15 2. Supplementing (reduced) earnings-related benefits Done in many countries, e.g. USA (401k): supplementary employer sponsored systems with incentive matching, partly in response to costs and risks of DB and increased labor mobility Austria, Italy & Korea: reformed severance pay (employer sponsored) and transition to pre-funding of benefits Germany (Riester pensions): Means tested contribution match, additional subsidy for all for number of children Czech Republic: fiscal flat-rate match Likely to gain importance as public generosity is reduced, governments and employer less willing to bear demographic and interest rate risks and consumption tax incentive questioned for reasons of effectiveness and equity (distributional outcomes) Key questions are effectiveness in producing net savings increases in proportion to fiscal costs 15

16 3. Coverage expansion within the Done in a few countries, e.g. mandated SI scheme o Mexico (quota social, 1 peso inflation adjusted): budget financed o Columbia (RAIS and FSP, to support low income workers): contribution financed o Peru (for mandated schemes - SNP and SPP and new voluntary scheme [Welfare Pension Scheme] to encourage workers and owners in micro-enterprises]: budget financed o Korea NPS for fishermen and farmers (in DB program) o Chile, young workers; [auto-enrolement for self-employed] Key aspect is to make mandated scheme more attractive to lower income groups in countries with moderate contribution density Utilize both ex-ante and ex-post matching design Experience/experiments have been in environments characterized by high levels of labor informality and large low income groups Potential likely to be greatest when there is large informal middle income population 16

17 4. Coverage expansion outside mandated SI schemes universal approach Done in very few countries, such as India (informal sector workers joining national defined contribution New Pension Scheme (NPS) established as substitute for fiscally unsustainable civil servants scheme): o Seeks to leverage record keeping and investment infrastructure created for reformed civil servants scheme o Use of aggregation agents and simplified account structure with lower fees to reach low income informal sector o annual match of contribution of 1,000 RS (US$ 25) for all who enroll no means test Relevant for low coverage countries with low income population outside the formal labor market where matching will not undermine formalization (in the short run) Key questions: Efficacy of the use of aggregators and points of presence to extend access into rural areas Vision of future joint system and transition toward; exit strategy 17

18 5. Coverage expansion outside mandated SI schemes sector/group specific approach Done in few countries, e.g. China (rural as of 2009, urban as of 2011) Tunisia, Capo Verde (plans for fisherman, etc) Korea (farmer and fishing sector) Peru SME micro enterprise Dealing with self-employed type individuals Key questions: Vision of future joint system and transition toward/ exit strategy Impact on mandated system/level of formality 18

19 III. Country Lessons across Income Levels: Coverage Expansion Participation From HICs - Effective but mostly small (5-10% increase compared to no match) with more effects when linked with nudging (e.g. auto-enrolement) Examples of higher effects (Germany), due to simplification and advocacy/marketing? Very successful in China with 358 million rural workers as of end-2011 and envisaged full coverage (of some 500 million) in

20 Coverage Expansion (2/2) Increase in scheme retirement saving o Typically very modest in size o Design matters (matching rate(s), threshold and structure) o Importance of decision architecture (defaults for rate levels, scheduling future increases now, etc) Non-decrease in other savings o Mixed results for US o Puzzles for UK pilot (rate up, wealth constant) o Claim for no substitution in Germany 20

21 2. Informality Matching incentives for system participation: No evidence of informality reduction for schemes in Mexico, Colombia and Peru In Chile matching for young workers has attracted participation but not clear about net effect In China main success in participation but those where strictly speaking not informal before 21

22 3. Fiscal Effectives Compared to ex-post subsidies Demogrant/citizen income Minimum pension guarantee Minimum income guarantee Accumulated budget expenditure compared to amount of additional saving created The stock of savings created by the scheme (net after administrative costs and before any disbursement is taking place) assuming that no saving substitution is taking place (i.e. a gross concept) The additional stock of national saving created (i.e. additional national capital stock) that excludes budgetary dis-saving induced by the expenditure 22 (i.e. a net concept)

23 V. Policy Conclusions and Next Steps 1. International experience that has been mostly collected from HICs suggests that MC are able to a. attract some but limited system participation b. provide at best moderate system saving effects for participating individuals, and c. offer quite likely insignificant wealth effects 23

24 Policy Conclusions (2/3) 2. The evidence from high income settings suggests that the nature and magnitude of effects are influenced by design and parameter choices 3. At least equally important are interventions to overcome behavioral limitations such as Information and awareness Financial literacy and skills Choice architecture (nudging et al) Social marketing, edutainment, advocacy and other non-cognitive interventions 24

25 Policy Conclusions (3/3) 4. There are many reasons why individuals stay away from formal systems. MC is not they way to overcome them they need to be dealt with directly through reforms 5. The quality information base for MC is slim and reduced to very few HICs. To pursue this policy approach requires rigorous evaluations at all country income levels 6. We lacking examples and framework for cost/benefit analysis for MCs 7. It is suggested useful to speculate about the role of MC in a future pension world 25