Early Life Schooling and Late Life Financial Literacy in The Wisconsin Longitudinal Study
Pamela Herd and Karen Holden
Prepared for presentation at the FLRC Conference, Washington, D.C.
FLRC
Conference, Washington, D.C.
November 18‐19, 2010
SSA FLFC 19-F-10003-5-01, University of Wisconsin
1
The research reported herein was performed pursuant p
p f
p
to a grant from the U.S. Social Security Administration (SSA) funded as part of the Financial Literacy Research Consortium. Consortium
The opinions and conclusions expressed are solely those The
opinions and conclusions expressed are solely those
of the author(s) and do not represent the opinions or policy of SSA or any agency of the Federal Government or of the of the University of Wisconsin System, including the Center for Financial Security.
SSA FLFC 19-F-10003-5-01, University of Wisconsin
2
Study Contributions
Study Contributions
• Focusing
Focusing on the relationship between on the relationship between ‘general
general skills’ early in life and financial literacy in late life rather than ‘specific
life rather than specific skills
skills’ training.
training
• Testing alternative measures of financial literacy.
literacy
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Changing Context of Retirement
Changing Context of Retirement
• Individualized retirement planning
Individualized retirement planning
– Shift from Defined Benefit to Defined Contribution
– Parallels in health care
Parallels in health care
• The rising importance of financial literacy skills
– Parallels in health care (health literacy)
.
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Specific vs. General Skills: Interventions
• Current
Current research emphasis on specific kinds of research emphasis on specific kinds of
‘educational’ intervention—specific skills
• But what about general skills? How does early lif
life schooling matter? Might these more h li
? Mi h h
general skills matter in the ability to adapt in changing contexts?
h i
?
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Measures of Financial Literacy
Measures of Financial Literacy
• Existing measures include:
Existing measures include:
– Knowledge (ability to calculate compound interest)
– Behavior based measures (i.e. debt levels, participation in retirement plans)
participation in retirement plans) • Proposed Measure: Knowledge of individual financial resources
financial resources
– Captures both knowledge and behavior
– Usefulness may vary based on age of respondents
Usefulness may vary based on age of respondents
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Research Question
Research Question
• How
How do early life schooling and general do early life schooling and general
cognition measures correlate with knowledge of individual financial resources in late life?
of individual financial resources in late life?
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Data: Wisconsin Longitudinal Study
Data: Wisconsin Longitudinal Study
• 1
1 in 3 Wisconsin high school graduates from in 3 Wisconsin high school graduates from
the class of 1957
• We employ the 1957 and 2004 data.
We employ the 1957 and 2004 data
• Sample loss due to death, attrition, and non‐
response.
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Data: Wisconsin Longitudinal Study
Data: Wisconsin Longitudinal Study
• Outcome Measures (2004):
Outcome Measures (2004):
– Ability to answer values on all asset questions (percent of assets held for which exact amount
(percent of assets held for which exact amount was given).
– Ability to provide value of retirement accounts Ability to provide value of retirement accounts
that accumulate a balance including IRAs, 401‐ks, etc.
– Ability to state value of checking account.
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Data: Wisconsin Longitudinal Study
Data: Wisconsin Longitudinal Study
• Covariate Measures (1957):
Covariate Measures (1957):
– Sex
– Age
– Parental SES
– Educational Attainment (updated in recent Ed ti
l Att i
t( d t di
t
surveys)
– Coursework (English, Algebra, Trig/Physics)
Coursework (English Algebra Trig/Physics)
– IQ (3 splines)
– High School Rank
Hi h S h l R k
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Methods
• OLS
OLS and Logit/Probit Models
and Logit/Probit Models
• Separate models for those with and without college degrees
college degrees.
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Findings
SSASSA
FLFCFLFC
19-F-10003-5-01,
19-F-10003-5-01,
University
University
of Wisconsin
of Wisconsin
12
Results:
Assets Summary
Results: Assets Summary
0.1
0
1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
‐0.01
College
No College
NS
NS
NS
NS
SSA FLFC 19-F-10003-5-01, University of Wisconsin
NS
Results: Assets Summary
Results: Assets Summary
• IQ
IQ and high school rank matter for those and high school rank matter for those
without a college degree. Up to 100, each 10 points is associated with a 3% increase asset
points is associated with a 3% increase asset knowledge.
• Coursework (Trig/Physics/Algebra) matter for Coursework (Trig/Physics/Algebra) matter for
those with a college degree. But these effects are smaller than the IQ effects for those
are smaller than the IQ effects for those without a college degree, but closer to the rank effect. effect
‘rank’
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Results: Private Pensions
Results: Private Pensions
02
0.2
0.15
0.1
College
NS
N C ll
No College
0.05
NS
0
Male
‐0.05
IQ Bottom
IQ Middle
Additional Schooling
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Results: Private Pension
Results: Private Pension
• For
For those without a college degree, IQ those without a college degree IQ
matters. (up to 120, each 10 points is associated with a 4% increase in the
associated with a 4% increase in the probability of knowing the value).
• For those with a college degree it
For those with a college degree it’ss additional additional
schooling that matters. (2% for each additional level)
additional level).
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Results: Checking Accounts
Results: Checking Accounts
0.12
0.1
0 08
0.08
0.06
Men
IQ Middl
IQ Middle
0.04
0.02
NS
0
College
No College
‐0.02
0.02
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Results: Checking Account
Results: Checking Account
• For
For those without a college degree, IQ those without a college degree IQ
matters. 10 points increase probability by 3 percent for middle IQ group
percent for middle IQ group. • For those with a college degree, algebra coursework is weakly related
coursework is weakly related. SSA FLFC 19-F-10003-5-01, University of Wisconsin
Sensitivity Analyses
Sensitivity Analyses
• Comparing do not knows to refusals
Comparing do not knows to refusals
– IQ
– Gender
– Underestimated effects
• Varying the course content measures
V i th
t t
• Probabilities instead of odds ratios
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Discussion
• Importance
Importance of early life cognition and of early life cognition and
educational attainment
• GENDER!
• Need more tests of these outcome measures (i
(i.e. utilize bracketing)
ili b k i )
SSA FLFC 19-F-10003-5-01, University of Wisconsin
Full contact info: Pamela Herd, 1080 Observatory Drive, Madison, WI 53706, email: pherd@lafollette.wisc.edu.
http://cfs.wisc.edu/
21