What is it with What is it with compulsory saving?compulsory saving?

Michael Littlewood

Retirement Policy and Research CentreRetirement Policy and Research Centre

A dAgenda

• Is New Zealand Superannuation sustainable?Is New Zealand Superannuation sustainable?• Do we need the New Zealand Superannuation

Fund?• Do we need KiwiSaver, let alone compulsion?• What can governments actually do?g y• What should New Zealand really do next?

NZS - can we afford it?• Population aged 65+ will more than double• Cost of NZS will more than double• Dependency ratio will worsen, but:

Total (as % GDP):32% i 2006- 32% in 2006

- 37% in 2050

NZS (net cost):- 3.4% in 2010- 7.1% in 2050

Source: Retirement Commissioner 2007

NZS – the Treasury’s LTFM 2000-2007• Demography drives expected cost• GDP drives ‘affordability’• 2009 estimates lowest in real terms

10%Net cost of New Zealand Superannuation - 2004 to 2100% GDP

V i 2 (2000)

7%

8%

9%

10%

Version 4 (2004)

Version 3 (2003)

Version 2 (2000)Version 1 (2000)

Version 5 (2005)Version 6 (2006)

4%

5%

6%

( )

Version 7 (2009)

1%

2%

3%

0%

2004

2007

2010

2013

2016

2019

2022

2025

2028

2031

2034

2037

2040

2043

2046

2049

2052

2055

2058

2061

2064

2067

2070

2073

2076

2079

2082

2085

2088

2091

2094

2097

2100

What about other countries? EU1525.0

Belgium

% GDP

20.0

BelgiumDenmarkGermanyGreeceSpainFrance

15.0

FranceIrelandItalyLuxembourgNetherlandsAustria

10.0

AustriaPortugalFinlandSwedenUK

5.0

Source: European Commision 2006

0.0

2004 2010 2015 2020 2025 2030 2040 2050

What about other countries? - 2

14.0

% GDP

10.0

12.0

EU25

6 0

8.0

4.0

6.0

New Zealand

0.0

2.0

2004 2010 2015 2020 2025 2030 2040 2050

Australia

Sources: European Commision, LTFM (net), Intergenerational Report 2010

2004 2010 2015 2020 2025 2030 2040 2050

Does the NZSF change things?

• The cost of NZS is not changed by $1• The incidence of cost is changed slightly:• The incidence of cost is changed slightly:

– Contributions holiday until 2018/19– By 2050, net outgo reduced by 0.6% of GDP (-8%)

• Exposure to risk• ‘Total accounting context’: borrowing to invest g g• Cost taxpayers $2.6 bn at 30 June 2009• Example of ‘cookie jar economics’p j• Purports to ‘lock-in’ NZS benefit design &

Budget process

We should agree what to pay for before how

What do we know?at do e o

• Public reviews – 1992, 1997, 2003, 2007 (2010 d )(2010 underway)

• New Zealanders’ responsesS i• Some questions:– Do tax incentives increase saving?

Might compulsion help?– Might compulsion help?– Can governments really change behaviour?– Does domestic saving matter?g– How do we resolve this apparent dilemma?

Let’s look at some evidenceLet s look at some evidence

www.PensionReforms.com

Can governments actually change things?

• International evidence is ‘no’• For example 48 country study: 1980–2004• For example, 48 country study: 1980 2004• $1 in pension saving adds 0-20 cents to

national savingg• Ignores cost of incentives and sub-optimal

investment decisions• Small “improvement” with maturity• “Reforming countries” don’t seem to be g

different

Pensions and Saving: New International Panel Data Evidence Bebczuk and Musalem (2006)

Can governments change things? - 2

• More evidence that answer is ‘no’S t t d 1970 t 2000• Seven country study – 1970 to 2000

• Voluntary pension savings largely not ‘new’ money money

• “We found substantial evidence that pension saving substitutes for other forms of private saving substitutes for other forms of private saving.”

Pension Reform and SavingBosworth and Burtless (2004)

Do tax incentives work? - 1

• Evidence that answer is ‘no’• No evidence that they increase savings• No evidence that they increase savings• … or ‘saving’ • They are regressive• They are regressive• … and increase the taxes of all• They also distort behaviourey a so d sto t be a ou

Current taxation of qualified pension plans: has the time come?

Munnell (1992)

Do tax incentives work? - 2

• Incentives change behaviour• Direct incentives probably don’t increase saving• “… between 0 and 30 percent of 401(k) balances

ddi i i i ”represent net additions to private saving”• Ignores direct/indirect costs of incentives

The Effects of 401(k) Plans on Household WealthEngen and Gale (2000)Engen and Gale (2000)

Does compulsion work?A 13 t i f L ti A i• A 13 country review of Latin America

• 11 lessons including:– Growing ‘informality’ of labour force– Growing informality of labour force– ‘Ownership’ doesn’t solve evasion– Suppliers tend to concentrate to a few– Competition doesn’t control costs– Market doesn’t solve mortality issues– Effect on national saving is uncertainEffect on national saving is uncertain– Large, regressive, long-tail costs in transition– May have made markets more liquid (but may not)– Investment risk adds to social risk

Reassessing Pension Reform in Chile and Other Countries in gLatin America

Meso-Lago (2002)

What drives saving?

• Higher output growth boosts saving• “Fiscal consolidation” linked with increased saving• Fiscal consolidation linked with increased saving• Private credit increases tend to reduce saving• Ageing populations reduce savingg g p p g• Better ‘terms of trade’ tends to increase saving• Saving behaviour may not be affected by returnsg y y• Increased credit may mean firms invest more• Higher cost of capital associated with lower

iinvestment

World Economic Outlook 2005World Economic Outlook, 2005International Monetary Fund

Higher savings = growth?

• More savings matter for ‘poor’ rather than ‘rich’ countriescountries

• Review of 118 countries over 1960-2000• Open capital markets disrupt theories based on • Open capital markets disrupt theories based on

closed economies• Local savings matter for innovation in ‘poor’ • Local savings matter for innovation in poor

countries – not significant for ‘rich’

When Does Domestic Saving Matter for Economic Growth? Aghion, Comin & Howitt (2006)

Are ‘household saving’ numbers helpful?e ouse o d sa g u be s e p u

• For ‘retirement saving’, stocks matter not flows“N thi b t ti t i d • “Nothing about retirement saving adequacy can be inferred from” household ‘saving’

• Possible adjustments could convert net national • Possible adjustments could convert net national saving from 2.1% to 13%

• Correcting for inflation removes “the so widely g ycited downward trend in private saving”.

• The ‘look around’ test is more usefulO C ll h h ld b• OECD will not now use our household numbers

Saving in New Zealand: measurement and trendsClaus & Scobie (2002)

New Zealanders behaving badly? - 1e ea a de s be a g bad y

• A look at available evidence• Review concludes that about one third are not • Review concludes that about one third are not

saving ‘enough’• Conservative assumptionsCo se at e assu pt o s• Data could be better• So, the problem is ……. ?, p

Are Kiwis Saving Enough for Retirement? Preliminary evidence from SOFIEfrom SOFIE

Trinh Le, Grant Scobie and John Gibson (2007)

New Zealanders behaving badly? - 2e ea a de s be a g bad y

• Treasury: A "least regrets" approach today seems defensible (2007)defensible (2007)

• Report: NZrs getting richer despite ‘spending more than they earn’

• ... “and not just because of housing”• “reasons that have been used to justify pro-saving

li i l k i d i i ”policies lack economic underpinnings.”• “Pro-saving policies are more likely to be regrettable

than not ”than not.• So, the problem is ……. ?

Does New Zealand have a household saving crisis?

Trinh Le, NZIER (2007)

New Zealanders behaving badly? - 3

• Compares SoFIE (individual) data for 2004 and 2006S t ‘ t’ d ‘t it ’ ff t• Separates ‘permanent’ and ‘transitory’ effects

• Measurement issues• Median saving rate: 16% p a of real incomes (5% • Median saving rate: 16% p.a. of real incomes (5%

ex-property) • Wide distribution – data may improveWide distribution data may improve• So, the problem is ……. ?

Saving Rates of New Zealanders: A Net Wealth Approach

Grant Scobie & Katherine Henderson (2009)

How are New Zealand’s old faring now?

• Ground-breaking work by MSD – ‘Economic Living Standards Index’ (ELSI)( )

• We must have been doing some things right• The old (65+) have the smallest levels of ‘hardship’:

only 8% have any at all• Unrelated to ‘financial assets’ – 59% have $25,000

or less or less • Owning a debt-free home is important• So the problem is ?• So, the problem is ……. ?

New Zealand Living Standards 2004Ministry of Social Development (2006)

More on New Zealand’s old• 2008 Living Standards Survey• Non-income measures of hardship• Supplements usual income-based measures• “older New Zealanders (65+) have low hardship

t (4%)” ( f EU25 14%)rates (4%)” (cf. EU25: 14%)• whole population: 13%; children 19%; sole

parents:39%; beneficiaries: 51%parents:39%; beneficiaries: 51%• So, the problem is ……. ?

Non-income measures of material wellbeing and hardship: first results from the 2008 New Zealand Living Standards Survey with international comparisons

Bryan Perry (2010)

SoFIE’s symmetry

• Eight year longitudinal survey• Substantial population sample• Substantial population sample• Started 2002 – first tranche of financial data

for 2003/2004for 2003/2004• Subsequent financial data 2006, 2008 & 2010• Straddles KiwiSaver’s introduction• Straddles KiwiSaver s introduction• We might be able to see impact• Australia’s equivalent is HILDA• Australia s equivalent is HILDA

HILDA (Aust) vs. SoFIE (NZ): 2006( ust) s So ( ) 006

As % net assets HILDA SoFIE

H + th 50 3% 46 2%House + other prop 50.3% 46.2%

Pensions/super 18.6% 2.1%

Businesses/farms 9.3% 22.2%Businesses/farms 9.3% 22.2%

Shares, funds etc 7.7% 9.3%

Bank accounts 4.7% 4.6%

Vehicles 3.9% 2.7%

Other assets 5.7% 13.9%

Liabilities (% gross) 14.4% 13.9%

HILDA vs. SoFIE: 2006s So 006

• Governments can influence aspects of behaviour• Probably have limited impact overall• Convertible to cash in 2006:

– Australians: 54.5% of net assets– New Zealanders: 48.5%

Debt abo t the same• Debt about the same:– Australians: 14.4% of gross assets– New Zealanders: 13 9% New Zealanders: 13.9%

More from HILDA: 2005-2009

• ‘Financial’ savings tracked (excludes home, lifestyle assets) by individuals) y

• Inflation-adjusted, ‘active’ saving• Savings (p.a.)g (p )

– All adult Australians: $300 p.a. (1.3% disposable income)

– Age 45-54: $2,260 p.a.– Age 65-74: -$110 p.a.– Top quintile earners: -$25,710 to +$39,120 (med. $8,060)

“With t SG t ib ti th t i l A t li • “Without SG contributions, the typical Australian would have spent $210 more than they received.”

Saving Tomorrow – the saving and spending patterns of AustraliansNATSEM, University of Canberra (2010)

New Zealanders behaving badly? - 4• Report compares SoFIE 2004 and 2006• Wealth at both years compared - ‘stocks’ measure

A i d l ( )• Average saving per adult (over two years):– Total: $29,900– Excluding housing gains: $12 600 (18% gross incomes)Excluding housing gains: $12,600 (18% gross incomes)– Excluding property & durables: $9,700

• “…passive saving from the property boom did not crowd out other forms of saving”

• Concerns on lack of saving “strongly overstated”S th bl i ?• So, the problem is ……. ?

Household Wealth and Saving in New Zealand: Evidence from Household Wealth and Saving in New Zealand: Evidence from the Longitudinal Survey of Family, Income and Employment

Trinh Le, John Gibson and Steven Stillman (2010)

Do we need KiwiSaver?Do we need KiwiSaver?

• New Zealanders are bad savers (?)( )• Australia shows the way (?)• Future retirees won’t have enough to live on (?)Future retirees won t have enough to live on (?)• We need to finance an ageing population (?)• New Zealand lacks sufficient local capital (?)New Zealand lacks sufficient local capital (?)• Fewer New Zealanders own their homes (?)• ‘Behavioural economics’ can help savers make • Behavioural economics can help savers make

‘appropriate’ saving decisions (?)

KiwiSaver – interim judgement

• Some positive aspects• Workplace participation has increased• Workplace participation has increased• But … the ‘problem’ not defined• Founded on questionable assumptionsq p• Next-to-no research – no debate• Rushed introduction; imperfect implementationp p• Not designed for employers• Tax breaks & housing subsidies unjustified• Introduces unnecessary public policy risks• Probably won’t ‘work’ – definition?

KiwiSaver – ‘Not achieved’

OECD measures ‘good’ regulation as:– Serving clearly identified goals– Having a sound legal & empirical basis– Producing benefits greater than costs– Minimising costs & market distortionsMinimising costs & market distortions– Promoting innovation through market incentives & goal-

based approachesB i l d ti l– Being clear and practical

– Being consistent with other regulations & policies– Being compatible with competition, trade & investment-g p p ,

facilitating principles

OECD Guiding Principles For Regulatory Quality And Performance OECD Guiding Principles For Regulatory Quality And Performance (2005)

Behavioural responsese a ou a espo ses

• Tax treatment is now complex and lacks:– Transparency– Logic– but costs more to administer– …but costs more to administer– …and the boundaries are constantly tested

• Amounts in superannuation will rise ….Amounts in superannuation will rise ….• …. but not necessarily ‘saving’• Tax planning is re-emergingTax planning is re emerging• “Social assistance integrity”?

What can governments really do?• Little influence over saving levels (except their

own)• Growth matters (almost above all else)• NZS design a strong influence on behaviour• Good data matter• Education matters

Di l tt ( d th l t i ) • Disclosure matters (and the regulatory regime) • Tax matters – TTE the ‘gold standard’• Everything else should be left to:• Everything else should be left to:

– Employers– EmployeesEmployees– Other individuals

What do we need to fix/discuss?What do we need to fix/discuss?

• Let’s start a proper, evidence-based debate on:– NZS benefit design from 2030 onwards– The NZ Superannuation Fund– KiwiSaverKiwiSaver– Tax incentives/ compulsory private provision– Tax regime for ‘collective saving vehicles’– Interface with income-tested benefits– Disclosure regime

• Depends on better information than we have• Depends on better information than we have• Requires a more robust framework than we have

been used to• Politicians can participate but can’t lead debate

A thought:“There is no greater tyranny than to force a man to pay for what he does not force a man to pay for what he does not want because you think it will be good for him.”for him.

Robert Heinlein