How to solve social security s problems in the

How to solve social security’s problems in the US By Estelle James 1

Some key numbers • 2005 -Payroll taxes >ss benefits so OASDI trust fund runs surplus that is lent to treasury for iou’s • Trust fund now holds $1. 6 trillion in iou’s • 2018—benefits > payroll taxes, so treasury begins to redeems iou’s • 2042—trust fund is empty, either pay 72% of scheduled benefits or raise tax to 19% (from 12. 4) • 75 -year unfunded liability=$3. 7 trillion • 2% permanent increase in payroll taxes might cover this deficit, if started today • Medicare is a much bigger problem 2

3 approaches (and where do individual accounts fit in? ) • Do nothing until 2042—at which point we need large benefit cut or tax increase (imprudent) • Start making gradual parametric changes soon – Raise revenues—payroll tax rate, ceiling, other – Cut growth in benefits or raise retirement age – This implies trust fund will be larger, last to 2080 • Start pre-funding—accumulate large funds that will reduce the unfunded liability, last after 2080 • Note that parametric changes also imply build-up of trust fund--so we really have 1 option but two key questions—how much & who manages funds? • Basic argument for IA’s: large pre-funding is 3 important & private management of funds is best

Criteria for evaluating alternative systems • Should be financially sustainable • Good for workers—reasonable benefit, not too risky or costly to administer • Equitable—tied to contributions, except for groups that we want to subsidize • Good for broader economy— – Keep payroll tax low, employment high – increase national saving to raise productivity • My bottom line: pre-funding through IA’s helps, if done right, but parametric changes also needed 4

Reasons for large pre-funding • Pre-funding makes system more sustainable, less sensitive to demographic change • Earns investment return, requires lower payroll tax —good for labor market • Pre-funding can help increase national saving, and financial market development--therefore productivity and growth • Avoids passing large debt to our children • (But, by itself, won’t bring system back to balance) 5

Problems with public management • Under public management (trust fund) treasury gets exclusive access to funds, iou’s don’t count in public deficit—may increase govt borrowing so treasury has to tax more, spend less later • If invest in stock market leads to: – conflict of interest between govt as regulator & investor – political lobbying for inclusion, exclusion • In other countries public funds get lower rate of return, political manipulation, misallocated capital 6

Private investment also entails problems • Financial market risk, uninformed investors, high administrative costs, accumulation may be used up too quickly • But with careful design and regulation by govt these problems can be resolved—checks and balances • Devil is in the details, in answers to key design issues. I will outline issues and choices made by other countries. 7

1. What is best mix of funding and PAYG? • Chile—almost 100% of system is in funded IA’s, except for safety net—MPG (10 -15%) • Sweden— 2. 5% tax goes into IA’s, 16% goes into PAYG, safety net financed by general revenues; IA’s only 15% of total payroll tax but expected to be 30% of total benefit • Eastern & Central Europe, Australia, Switzerland, Netherlands—private funded pillar 50% of total • Mixture of PAYG & funding diversifies risks • Small IA is under consideration in US--important to realize we will retain traditional benefit 8

2. If private, who chooses investment managers? • Workers—Latin America, Eastern & Central Europe (retail market-may lead to high costs) • Employers—Switzerland, Netherlands, Denmark, Australia, Hong Kong – Long tradition of employer-sponsored plans, government recently made them mandatory – Group plans cheaper to administer (institutional market) – But if defined contribution (DC) worker bears risk and will demand choice of investment manager & strategy • Thrift Saving Plan in US, Bolivia, Kosovo— competitive bidding process chooses small # of 9 asset managers—enables workers to get lower cost

3. How to keep administrative costs low • Chile—administrative costs high at start-up but < 1% of assets now (< US mutual funds) • Higher in new plans in Latin A. , E. Europe • Costs much lower in Thrift Saving Plan, large employer plans. Why? What can we learn? – Aggregating small accounts increases bargaining power, reduces marketing costs. Use competitive bidding process to choose small number of asset managers (limits workers choice) – Use passive investing (indexing to benchmark like S&P 500 or Russell’s 3000)—(didn’t exist in Chile) – Collect contributions through tax system, centralize record-keeping (less money, more time) 10

4. How to reduce risk and provide safety net to low earners • Every country with IA has minimum pension or flat benefit that sets floor – MPG in most Latin America, Eastern Europe – Flat or almost flat benefit in Western Europe, Australia – Switzerland also sets floor on interest rate paid by pension funds & insurance companies • Require broad diversification of investments(int’l) • Guarantees often required. But: – Private market can provide guarantees but difficult in long run (costly, uncertain, hard to evaluate price) – But public guarantees often cost more than expected, moral hazard problems 11

5. How to handle payouts • How can we be sure that worker won’t spend all his money before he dies? Annuities guarantee workers life-long income. Can private sector handle this? Will annuities be offered on good terms? Will workers buy? • Chile has had IA system since 1982, many workers have retired, 2/3 annuitized—government requires annuities or gradual withdrawals. • Annuities must be price-indexed, joint for married men--inflation insurance, protection for widows • Private sector can handle payout stage, but only if government sets careful rules of the game 12

6. How to protect women • Women work less, earn less, live longer, and if married their husband are older, so they become widows with low incomes. Very old women are pockets of poverty in many countries. • Women are the biggest gainers in lifetime income from the pension reform in Latin America. Why? – Minimum pension guarantee or flat benefit helps low earners – Husbands required to purchase joint pension--survivors benefits financed by husbands (in US by taxpayers) – Widows keep their own pension plus joint pension (in US widows must choose between the two; many women who work & contribute get no additional benefit) 13

7. How to handle the transition • Chile, Latin America, Central & Eastern Europe— used carve-out from existing payroll tax that was high at start. Therefore transition financing gap (part of tax diverted to IA while still had to pay benefits) – Most downsized current benefits – Chile accumulated fiscal surplus before reform • Australia, Netherlands, Switzerland, Denmark, Hong Kong—add-on, no transition cost problem • US—treasury deficit, trust fund surplus til 2018; to increase national saving we shouldn’t finance transition by increasing public deficit (otherwise higher personal saving offset by public dissaving) 14

My preferred plan • Start IA with mixed add-on + carve-out—more money to keep benefit high, transition cost low • Carve-out progressive—higher for low earners • Add minimum pension per year worked—to offset labor & financial market risk • Index retirement age to longevity (automatic raise) • Don’t price-index traditional benefit—keep it rising with wages for successive cohorts • Finance transition by higher ceiling on payroll tax or surtax on all income—not debt finance • Use Thrift Saving Plan model to keep costs low • Spread payouts over lifetime, joint for couples 15

Conclusion • Mixture of PAYG DB and funded IA would make social security sustainable, protect workers, and improve economy—at least cost • . But if we don’t start now, pre-funding won’t be an option. • The devil is in the details. IA’s can make us better or worse off, depending on design. So we should focus on “what is best design? ” Experience of other countries suggests design options. • Ultimately, regardless of structure, we will have to make value judgment between reducing benefits vs. raising payroll taxes (contribution to IA is only 16 quasi-tax and gives us better terms of trade).