Innovations in Savings: Policy Implications from g y p Experiments in

Innovations in Savings:
g Policy
y Implications
p
from
Experiments in the Field
$aveNYC: Applying Behavioral Economics
Principles to Encourage Savings
$aveNYC Research
Purpose: Test a scalable, national model to increase savings among
low- income individuals.
Intensive Research Component funded by Ford Foundation. Key
research questions:
• Will very low-income people save when given the opportunity and
support?
• Does starting to save fundamentally impact both long-term aspirations and
shorter-term behaviors and habits
• Can incentivized savings programs be delivered at scale?
Behaviorally-informed design
Limit Choices
Direct depositing at least $200
from tax refund
• 31% off people
l who
h rejected
j t d th
the
account reported that not wanting
an account at that particular
institution influenced their decision
not to participate.
50% match if participants
save for 1 year
Hassle Factors
Account opening at VITA site
• 67% reported that ease of opening
$
$aveNYC
NYC accountt iinfluenced
fl
d
decision to open “A LOT” – only
64% said the same about the
match!
Behaviorally-informed Design
Mental Accounting
“($aveNYC)
“($
NYC) iis money I
know I have and cannot
touch.”
Loss Aversion
“I would never have been
able to do it by myself”
myself
• 59% participated because funds would
be hard to access
• Drawback: 39% did NOT participate for
this same reason
• All participants intend to save for the
full year; think of this money as
savings
savings.
• 79% would withdraw from general
g for emergency
g
y spending;
p
g; only
y
savings
47% would withdraw from their
$aveNYC account
•
people
p with $50 match closed
40% of p
Potential implications for policy
•
Match incentives can be used to induce even very low income people
to save.
•
Tax refunds
T
f d presentt a unique
i
opportunity
t it to
t transform
t
f
intentions
i t ti
into
i t
savings.
– Tax refund was 21% of participants’ annual income.
– Even without a savings match
match, simple savings mechanisms such as
savings bonds and on-form account opening could have strong impact.
•
Flexible savings
g uses are key.
y
– The vast majority of participants did not have a specific asset goal. Asset
building policy should reflect the need for flexible savings.
Pilot Replicated through Social Innovation Fund
$aveUSA
•
Five cities selected to implement:
• New York
• Newark
• San Antonio
• Tulsa
• Savannah
•
p
will test implications
p
for federal p
policy
y shift to
National Replication
medium-term savings supports at tax time
•
3 years of program implementation; 5-year total evaluation (random
sample assignment in 2 cities).
cities)
Initial Pilot Results:
given the right
g
Low income households can save when g
incentive and opportunity
$aveNYC Highlights
2008
2009
2010
Accounts Opened
177
1063
1435
Accounts funded by IRS
151
952
1335
$387
$381
$704
6%
9%
10%
10%
18%
18%
Average Savings
Participation Rate (eligible)
Participation Rate (direct deposit filers)
Early findings on program use and outcomes
•
2009 S
Survey
r e (Cohort 1
1, Wa
Wave
e 1)
•
Preliminary analysis while waiting for follow-up data
•
Goal is to understand program dynamics and inform future efforts
•
Outcomes
•
Account opening
•
Amount of initial contribution
•
Savings persistence
Predictors of Account Opening
•
•
•
•
Demographic Characteristics
• Married respondents 88% more likely
• Latinos slightly less likely than African-American respondents to
open an account
Economic Characteristics
• Refund amount matters. Those with refunds over $1,500 nine
times as likely to open an account than those with <$500
Financial Behavior
• Those who report spending less than peers twice as likely to open
as those who report spending more
Interesting non-findings
• Income does not seem to matter much
• No effect from experience of financial hardship in the past year
Predictors of Initial Deposit
•
•
•
•
•
Median contribution:
contrib tion $500
$500, Mean contrib
contribution:
tion $388
Demographic Characteristics
• Latinos and those of other race contribute more than African
American respondents
Economic Characteristics
• Refund amount matters. Recipients of $200-$500 contribute $175
more recipients of >$500 contribute $258 more
more,
• Higher income households contribute slightly more
• Those who feel low on cash at the end of the month contribute
less
Financial Behavior
• Those confident in money management contribute less
• Those who describe themselves as savers contribute less
Interesting non-findings
• No effect from experience of financial hardship in the past year
Program Behavior Patterns
― High
g balance savers most likely
y to save for a full yyear
80% of account-holders saved
for a full year
32% of filers who contributed
less than $200 closed their
account before receiving match,
compared to 16% of filers who
saved more than $200
Predictors of savings persistence
•
•
•
•
Demographic Characteristics
• US born
• Employed
• Education
• Age
Economic Characteristics
• Feel low on cash at the end of the month
• Inability to pay bills
Financial Behavior
• Confidence in money management
• Regular savings habit
Interesting
g non-findings
g
• Initial contribution amount
• Income
CCC - $aveNYC Research
•
Multi-year, multi-cohort evaluation of treatment and comparison group
•
Economic and social outcomes related to program participation
•
Follow-up survey for 2009 cohort in the field now
•
Follow-up survey for 2010 cohort to be fielded in the summer of 2011
•
Integration of administrative account data
42 Broadway, 8th Floor
New York
York, NY 10004
www.nyc.gov/ofe
CB#3452
Chapel Hill, NC 27599
cckey@unc edu
[email protected]
www.ccc.unc.edu