Transcript How the Income Tax Treatment of Saving and Social Security

401(k) Participant
Behavior in a Volatile
Economy
Prepared for the 14th Annual RRC Conference,
August 2, 2012
by
Barbara Butrica and
Karen Smith
1
Background
• DC pensions have become the dominant
employer-provided pension type over the
last several decades.
• Compared with DB pensions, DC pensions
force workers to take charge of their
retirement investments.
• Whether to participate
• How much to contribute
• What to invest in and how much
2
What happens to 401(k)
contributions in a volatile economy?
• Three recessions
• Early 1990s (oil price shock)
• Early 2000s (collapse dot-com bubble, and
9/11)
• Late 2000s (stock market crash, housing
collapse, and Great Recession)
• Volatility
• Stocks, housing, employment
3
Questions
• How do workers’ 401(k) contributions
respond to economic downturns?
• Do they stop contributing to DC pensions?
• Do they change their contribution amounts?
• Do they change their asset allocation with
fluctuations in the stock market?
• Do responses vary depending on the size
of the downturn?
4
Data
• 1996, 2001, 2004, 2008 SIPP
• Large representative samples of US households
• Includes demographic, economic, and job
characteristics
• Linked to Detailed Earnings Record (DER)
• Includes total earnings 1978-2010 and worker
contributions to DC plans 1990-2010
• Sample is person-year file of workers ages
20 to 69 from 1990 to 2010
• 905,381 person-year observations
• Over 168,000 individuals
5
Methods
• Examine participation of workers by year, age,
income (DC contribution>0)
• All workers
• Workers offered a DC pension (SIPP pension year)
• Examine contribution amount among participants
by year, age, income
• Examine asset allocation from SIPP self-reports
• Flows (among contributors)
• Assets (among account holders)
• Multivariate analysis of participation, contribution
amounts, and stock allocation
6
Methods
• Particular variable of interest is YEAR
• Interested in variation around recessions and
stock market crashes.
7
Participation
8
Participation rate among workers ages 20
to 69 increased dramatically over time.
.
9
Participation rate declined slightly
with the recent recessions.
.
Drop in participation
in 2001 and 2008
recessions
10
Participation rate increased for all
workers—particularly those ages 30 to 64.
.
50%
50-59
45%
40-49
40%
30-39
60-64
35%
Participation Rate
30%
65-69
25%
20-29
20%
15%
10%
5%
0%
Year
11
They continued to rise through the recessions
for workers ages 60 to 69.
.
50%
50-59
45%
40-49
40%
30-39
60-64
35%
Participation Rate
30%
65-69
25%
20-29
20%
15%
10%
5%
0%
Year
12
And fell sharply following both the 2001 and
2008 recessions for workers ages 20 to 39.
.
50%
50-59
45%
40-49
40%
30-39
60-64
35%
Participation Rate
30%
65-69
25%
20-29
20%
15%
10%
5%
0%
Year
13
The increase in participation rate seems to have
plateaued below 50% for workers.
.
48%
50%
1966-1970
45%
1961-1965
40%
1956-1960
35%
Participation Rate
1951-1955
30%
1946-1950
26%
25%
1941-1945
20%
1936-1940
1931-1935
15%
1926-1930
10%
1921-1925
5%
0%
20-29
30-39
40-49
50-59
Age
60-64
65-69
14
The increase in participation rates seems to
have plateaued below 50% for workers.
.
48%
50%
1966-1970
45%
1961-1965
40%
1956-1960
35%
Participation Rate
1951-1955
30%
1946-1950
26%
25%
1941-1945
20%
1936-1940
1931-1935
15%
1926-1930
10%
1921-1925
5%
0%
20-29
30-39
40-49
50-59
Age
60-64
65-69
15
The increase in participation rates seems to
have plateaued below 50% for workers.
.
48%
50%
1966-1970
45%
1961-1965
40%
1956-1960
35%
Participation Rate
1951-1955
30%
1946-1950
26%
25%
1941-1945
20%
1936-1940
1931-1935
15%
1926-1930
10%
1921-1925
5%
0%
20-29
30-39
40-49
50-59
Age
60-64
65-69
16
Limitation of data
• Do not know if worker is offered a DC plan
in the DER data.
• Do know DC offer at the SIPP pension
interview.
• Limit the sample to include only workers
offered a DC plan.
17
Given an DC offer, two-thirds of workers
participate
.
100%
All Workers
90%
Workers Offered Plan
80%
68%
Partcipation Rate
70%
67%
65%
59%
60%
55%
50%
40%
72%
71%
41%
38%
45%
47%
38%
30%
26%
22%
20%
10%
0%
All
20-29
30-39
40-49
Age
50-59
60-64
65-69
18
Over 40 percent of workers can’t participate
because they are not offered a DC plan
.
100%
90%
80%
37%
43%
70%
36%
55%
21%
18%
Not Offer
42%
62%
60%
50%
36%
19%
18%
Offer Do Not
Participate
20%
40%
20%
20%
16%
30%
42%
39%
46%
46%
38%
24%
23%
10%
Offer and
Participate
0%
All
20-29
30-39
40-49
Age
50-59
60-64
65-69
19
Participation rate among workers offered a DC
plan fell with the Great Recession at all ages.
.
80%
70%
68%
71%
67% 65%
65%
59%
60%
72%
67%
68%
65% 65%
56%
55%
51%
Partcipation Rate
50%
40%
2006
2009
30%
20%
10%
0%
All
20-29
30-39
40-49
Age
50-59
60-64
65-69
20
Contribution Amount
• Include only contributors ages 20 to 69.
21
Median contribution amount among
participants fell with the Great Recession.
$4,000
-9.6%
-9.2%
DC Contribution Amount (in 2011 dollars)
$3,500
-3.4%
-6.6%
$3,000
-5.4%
-6.1%
$2,500
2007
$2,000
2008
2009
-6.5%
$1,500
$1,000
$500
$0
All
20-29
30-39
40-49
Age
50-59
60-64
65-69
22
Median contribution amount among
participants fell with the Great Recession.
$4,000
-9.6%
-9.2%
DC Contribution Amount (in 2011 dollars)
$3,500
-3.4%
-6.6%
$3,000
-5.4%
-6.1%
$2,500
2007
$2,000
2008
2009
-6.5%
$1,500
$1,000
$500
$0
All
20-29
30-39
40-49
Age
50-59
60-64
65-69
23
Median contribution amounts among
participants fell with the Great Recession.
$4,000
-9.6%
-9.2%
DC Contribution Amount (in 2011 dollars)
$3,500
-3.4%
-6.6%
$3,000
-5.4%
-6.1%
$2,500
2007
$2,000
2008
2009
-6.5%
$1,500
$1,000
$500
$0
All
20-29
30-39
40-49
Age
50-59
60-64
65-69
24
Median contributions continued to fall
through 2010 for younger participants.
$4,000
-6.7%
-5.1%
DC Contribution Amount (in 2011 dollars)
$3,500
-4.7%
-8.2%
$3,000
-5.3%
-6.2%
$2,500
2007
$2,000
2008
2009
-12.2%
$1,500
2010
$1,000
$500
$0
All
20-29
30-39
40-49
Age
50-59
60-64
65-69
25
Median contributions continued to fall
through 2010 for younger participants.
$4,000
-6.7%
-5.1%
DC Contribution Amount (in 2011 dollars)
$3,500
-4.7%
-8.2%
$3,000
-5.3%
-6.2%
$2,500
2007
$2,000
2008
2009
-12.2%
$1,500
2010
$1,000
$500
$0
All
20-29
30-39
40-49
Age
50-59
60-64
65-69
26
Median contributions increased in 2010 for older
participants, but remain below 2007 amounts.
$4,000
-6.7%
-5.1%
DC Contribution Amount (in 2011 dollars)
$3,500
-4.7%
-8.2%
$3,000
-5.3%
-6.2%
$2,500
2007
$2,000
2008
2009
-12.2%
$1,500
2010
$1,000
$500
$0
All
20-29
30-39
40-49
Age
50-59
60-64
65-69
27
Change in DC Contribution Amount
(percent)
-1
-3
-4
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
Changes in median DC contributions
closely align with changes in GDP.
5
4
3
2
1
0
-2
GDP
Median Contribution
Year
28
Change in DC Contribution Amount
(percent)
-1
-3
-4
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
Changes in median DC contributions
closely align with changes in GDP.
5
4
3
2
1
0
-2
GDP
Median Contribution
Year
29
Multivariate Analysis
• Participation
• Logistic regression all workers
• Control for age, earnings, number of work years since age 20,
job change, job loss, SIPP panel, year
• Logistic regression workers offered DC plan
(duration of pension job at SIPP interview)
• Control for age, earnings, number of work years since age 20,
job change, job loss, SIPP panel, year
• Marital status, change in marital status, spouse employment
status, spouse DC contributions, home ownership, number of
dependents, having a baby, work limitations, health status,
whether the employer contribute to the plan, can borrow from
plan.
30
Probability of contributing to DC plan declined with the
2008 recession. (Unadjusted Beta for year)
1
0.9
Participation Beta
0.8
All Workers
Workers Offered Pension
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Year
31
Probability of contributing to DC plan declined with the
2008 recession. (Unadjusted Beta for year)
1
0.9
Participation Beta
0.8
All Workers
Workers Offered Pension
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Year
32
Multivariate Analysis
• Contributions
• OLS regression all participants
• Control for age, earnings, number of work years since age 20,
job change, job loss, SIPP panel, year
• OLS regression all participants from 1990-SIPP
• Control for age, earnings, number of work years since age 20,
job change, job loss, SIPP panel, year
• Marital status, change in marital status, spouse employment
status, spouse DC contributions, home ownership, number of
dependents, having a baby, work limitations, health status,
whether the employer contribute to the plan, can borrow from
plan.
33
Real DC contributions declined with the 2001 and 2008
recessions. (Unadjusted Beta for year)
600
All Participants 1990-2010
500
Participants 1990 to SIPP
400
300
200
100
0
-100
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
-200
1992
OLS parameter estimate for Year
700
Year
34
Real DC contributions declined with the 2001 and 2008
recessions. (Unadjusted Beta for year)
600
All Participants 1990-2010
500
Participants 1990 to SIPP
400
300
200
100
0
-100
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
-200
1992
OLS parameter estimate for Year
700
Year
35
What does this mean for
retirement saving?
• Simple simulation
• Baseline uses observed median DC contributions for
30-39-year-olds from 2007 to 2010.
• Assume 3 percent real growth on accumulations
• Assume workers increase annual real contributions by 1
percent each year after 2010.
• Alternate uses observed 2007 median DC contribution
for 30-39-year-olds.
• Assume 3 percent real growth on accumulations
• Assume workers increase annual real contributions by 1
percent each year after 2007.
• Only difference is the DC contributions for
2008-2010.
36
For a typical 30-year-old, simulated retirement
saving is 9.1 percent lower at age 62 because
of the recession.
Retirement Saving ($2011)
DC Contribution ($2011)
3,500
3,000
2,500
2,000
1,500
Baseline
1,000
Alternative Scenario
500
160,000
$142,141
140,000
Baseline
120,000
$130,233
Alternative Scenario
100,000
80,000
60,000
40,000
20,000
0
0
30
35
40
45 50
Age
55
60
30
35
40
45 50
Age
55
60
37
Account holders were less likely to invest in
stocks after the 2008 economic melt-down but
contribution allocations remained stable
Probability of Investing in
Equities
70%
60%
64%
59%
52%
50%
49%
50%
53%
40%
Contributions
30%
Balances
20%
10%
0%
2003
2006
Year
2009
38
Conclusions
• Less than half of workers participate in DC
plans.
• 40 percent don’t participate because their
employers don’t offer a plan.
• Median contributions are well below the
statutory limit.
• Workers don’t save adequately even in economic
booms. They save less in recessions.
39
Conclusions
• Workers do lower DC participation and contributions
during recessions.
• Lower contributions precede the recession.
• Lower participation lags behind the recession.
• The Great Recession will lower retirement saving for
the typical 30-year-old today by about 9 percent.
• The impact is greater for higher-income workers and for
workers farther from retirement.
• Some evidence that investors buy high and sell low—
locking in investment losses.
• Financial literacy could improve retirement saving
40