Transcript Trinh Le 2008 Intervention in household saving decisions

Intervention in household saving
decisions
Symposium on Retirement Income Policy
Retirement Policy and Research Centre
Trinh Le – 16 April 2008
Outline
► Household
saving behaviour before KiwiSaver
► Effects of KiwiSaver – initial evidence
► How will KiwiSaver affect New Zealanders’ future
decisions?
2
Was there a saving problem?
► KiwiSaver:
an interventionist policy to correct a
market failure
► The failure: before KiwiSaver kiwis did not save
► The average NZ household spent $1.15 for
every dollar earned (Finance Minister Michael Cullen, Budget 2007)
► True, see data from the Institutional Sector
Accounts
3
Household saving: aggregate evidence
Household saving as % of household disposable income
5
0
-5
-10
-15
-20
1987
1990
1993
1996
Source: Household Income and Outlay Accounts
1999
2002
2005
4
More on saving evidence
Household saving as % of household disposable income
10
0
-10
Household Income and Outlay Accounts
Household Economic Survey
-20
1984
1987
1990
1993
1996
1999
2002
2005
5
More on saving evidence
Household saving as % of household disposable income
100
50
0
Household Income and Outlay Accounts
Household Economic Survey
Reserve Bank of New Zealand
-50
1979
1982
1985
1988
1991
1994
1997
2000
2003
2006
Source: Trinh Le, Does New Zealand have a household saving crisis?
NZIER working paper 2007/01
6
Was there a saving problem?
► Different
data tell different stories
► NZ doesn’t have a saving problem as much as a
saving data problem
► The data that are used to support the claim that
NZ households are bad savers are unreliable
► Careful analysis shows little support for that
claim
7
Micro evidence
► Even
when error free, aggregate data on saving
not as helpful as one might think
► Aggregate saving tends to 0
 Some borrow
 Some save
 Some dissave
► Have
to look at micro evidence to judge whether
people are adequately prepared for retirement
8
Modelling adequacy of retirement saving
► Life
cycle model
► Objective: consumption smoothing
 Given their current income, assets, liabilities, age and
life expectancy at retirement, how much should a
single person/ couple save so that when they retire,
they will have a standard of living that is similar to
their current standard of living?
► Used
unit-record data from the Household
Savings Survey (2001)
► Found on average, people already saved more
than “prescribed” by the model
9
Saving adequacy
Prescribed Actual saving Actual saving
saving rate
rate % rate %:
%
durables as
consumption
Mean
Median
7.2
26.6
16.0
27.0
Sample: Couples born 1940-1949
10
Saving adequacy
Prescribed Actual saving Actual saving
saving rate
rate % rate %:
%
durables as
consumption
Mean
Median
7.2
26.6
21.4
16.0
27.0
22.7
Source: Grant Scobie, John Gibson & Trinh Le (2005)
Household wealth in New Zealand
11
Modelling adequacy of retirement saving
► Analysis
repeated on a larger, newer data
(Survey of Family, Income and Employment,
wave 2, conducted 2003-04)
► Found most were saving enough
► Under relatively conservation assumptions
► Significant proportions
 have saved so much they shouldn’t save more
 have such low incomes that they shouldn’t save given
that NZ Super is very high compared to their current
income
12
Bad savers?
% saving inadequately
Non-partnered individuals
Ages 45-54
18
Ages 55-64
9
Ages 45-54
Ages 55-64
26
13
Couples
Source: Trinh Le, Grant Scobie & John Gibson (2007)
Are kiwis saving enough for retirement?
13
‘Too good’ savers
% prescribed with
negative saving rate
Non-partnered individuals
Ages 45-54
58
Ages 55-64
63
Ages 45-54
32
Ages 55-64
37
Couples
Note: % of people who, given their current income and wealth,
have saved enough for retirement and shouldn’t save more
14
Inertia – behavioural economics
► A key
argument for KiwiSaver is that people are
myopic, they don’t plan for a distant future, they
often procrastinate saving until too late
► Auto-enrolment was believed to overcome the
power of inertia
► In fact, far more people have joined KiwiSaver
through direct enrolment than auto-enrolment
15
KiwiSaver enrolments
Auto enrolled
25%
Opted in
75%
Source: Inland Revenue, as at 29 Oct 07,
Total membership: 251,736
16
Does KiwiSaver mean more saving?
► As
of April 2008, over 500,000 kiwis have joined
KiwiSaver
► The first six-monthly evaluation report did not
evaluate the very question it should evaluate:
Does KiwiSaver increase saving?
► Gibson and Le (2008) made the first attempt and
found KiwiSaver a money-go-round game
17
Composition of KiwiSaver balances
Self (new saving)
19%
Taxpayers
51%
Self (reshuffle)
21%
Employers
9%
Note: Upper bound estimate of new saving
18
Composition of KiwiSaver balances
Self (new saving)
9%
Taxpayers
50%
Self (reshuffle)
32%
Employers
9%
Source: John Gibson & Trinh Le, How much new saving will KiwiSaver produce?
University of Waikato Economics Department working paper 2008/03
19
How much KiwiSaver saving is “new”?
► 9-19%
of KiwiSaver balances
► Not enough to cover deadweight costs of
taxation
 which Treasury conservatively estimates to be 20% of
the amount of taxes raised
► Let
alone administration and compliance costs
► Net of these costs, the impact of KiwiSaver on
national saving most likely negative
► These findings are consistent with…
20
International evidence
Engen & Sholz (1994)
Scheme
IRA
New saving
-14% to 2%
Engen, Gale & Scholz (1994)
IRA
4%
Attanasio & DeLeire (2002)
IRA
9%
Engen & Gale (2000)
401(k)
0–30%
Benjamin (2003)
401(k)
25%
Note: Excl. deadweight costs of taxation and administration and compliance costs
21
Back to macro data
► NZ’s
large current account deficit (CAD) is often
used as evidence of a (household) saving
problem
► Australia’s compulsory saving scheme often
used as “role model” for NZ
22
Household saving & Balance of payments
5
0
-5
-10
-15
-20
Household saving rate
Current account balance as % of GDP
1987
1990
1993
Source: Statistics New Zealand
1996
1999
2002
2005
23
Back to macro data
► Australia
had falling household saving rates and
widening CAD in the 1970s and 1980s
► A compulsory workplace saving scheme
(Superannuation Guarantee) was introduced in
1992
► Both household saving and CAD have worsened
since
► Trends in household saving and CAD in Australia
are very similar to those in NZ, despite a
compulsory saving scheme
24
Australia
20
Household saving rate
Current account balance as % of GDP
15
10
5
0
-5
1970
1980
Source: OECD Economic Outlook
1992: Super Guarantee
2000
2005
25
Back to macro data
► Similar
patterns for US, despite long established
saving schemes like Individual Retirement
Accounts (IRA) and 401(k)
26
USA
Household saving rate
Current account balance as % of GDP
10
5
0
-5
1970
1981: 401(k)
Source: OECD Economic Outlook
1990
2000
2005
27
Summary
► Household
saving behaviour before KiwiSaver
 Retirement saving: most saved enough
 Aggregate saving: no reliable evidence of a saving
crisis
► Initial
effects of KiwiSaver
 Some impact on household saving
 Negative impact on national saving
► Expected
future effects of KiwiSaver
28