Download - Promise & Potential of U.S. Savings Bonds
© 2007, D2D Fund, Inc. 1
Promise & Potential of
U.S. Savings Bonds
Peter Tufano, Harvard Business School & D2D FundTimothy Flacke, D2D Fund
June 20, 2007
© 2007, D2D Fund, Inc. 2
Agenda
� Introductions
� LMI saving, tax time & the bucket problem
� Reinventing U.S. Savings Bonds
� Some context
� An experiment
� Findings
� Good news / bad news
� Suggestions
© 2007, D2D Fund, Inc. 3
Why Are We here?
� Care about saving, especially for Low toModerate Income (“LMI”)
� Why save?
� Shocks (emergencies)
� Strategic investments (e.g., home, education)
� Deferred consumption (e.g., retirement)
� How do we help low-income householdsto save?
© 2007, D2D Fund, Inc. 4
How Can We Help Families Save?
� Force them to save
� Make it hard not to save
� Make it easy to save
� Bribe them to save
� Social support for savings
� Make them excited to save
Social Security; UK Child Trust Program
Defaults and opt outs; bundling
SMaRT plan, Split Refunds and Savings
Tax advantages, matching funds
AmericaSaves
“Prize-linked” savings
© 2007, D2D Fund, Inc. 5
LMI Saving? Go Where the $$$ Are
� Tax Time combines:
� Savable funds (tax refunds)
� Topical tie in (finances top of mind)
� Intermediation (tax preparer typically involved,
possible advisor or salesperson)
© 2007, D2D Fund, Inc. 6
Tax Time: How Much Money?
� $220 billion (2003) in total refunds
� $103 billion for AGI < $40k
� $82 billion for AGI < $30k
� Average refund $1,482 (AGI < 30k)� 5% of annual income in single shot
� Median net financial assets only $2,000 (forbottom quintile by income)*
� Meaningful saving opportunity, especiallyrelative to current savings
* Source: Aizcorbe, Kennickell, and Moore (2003).
© 2007, D2D Fund, Inc. 7
How to Capture Refunds as Savings?
� Refund Splitting� direct portion of refund directly to savings
� where possible, offer new accounts during taxprep process
� Pilot tests TS 2004 - 2006� Demand as high as 13%
� Ultimate outcome: Form 8888
© 2007, D2D Fund, Inc. 8
But Saving Requires a “Bucket”
� Sizeable segment can’t get saving product:� Mainstream institutions
� disinterested or fearful (ChexSystems review)
� products aren’t inviting (min. balances, fees)
� often little physical presence in LMI communities
� Others (check cashers, payday lenders)� aren’t depositories / don’t offer saving
� Result: a pipe fromsavable refunds to...nowhere
© 2007, D2D Fund, Inc. 9
Where Can We Find a Bucket?
� Imagine a universalsavings vehicle
� Characteristics of thisvehicle?� small denomination� portable� competitive return� no principal risk� good liquidity...
� Product already exists:
U.S. Savings Bonds
© 2007, D2D Fund, Inc. 10
How Did We Get Here?
� 2005: Reinventing Savings Bonds
� Thought piece: how to reinvent bonds toserve LMI market
� Recommendations:
� Offer bonds to tax return filers
� Reinvigorate bond marketing
� Recognize bonds as part of financial lifecycle
� Clarify, expand emergency access to bonds
� Don’t discourage bonds w/ means testing
© 2007, D2D Fund, Inc. 11
Laying the Groundwork
� 2006: Just Keep my Money� Report on “pre-experiment” test of refund-
driven savings bond program
� test operational feasibility of selling bondsin commercial tax prep context
� trying to merge tax refund & bond orderprocess from point of view of IRS
� Indication of demand� but sample too small, too late in tax
season, product too limited (no gifting)� further study required
© 2007, D2D Fund, Inc. 12
This Year’s Agenda
Ran experiment to answer:
Will consumers buy bonds at tax time?
Preview: YES!� Experiment offered bonds to 8,000+ clients� 5.9% - 9.6% of clients offered, bought� 510 clients bought bonds for 881 people
� Substantial demand for bonds relative toother tax time saving products
© 2007, D2D Fund, Inc. 13
Context: An American Saving Crisis
© 2007, D2D Fund, Inc. 14
Beyond the Headlines
� In 2004, bottom 25% of U.S. householdsby net worth had:� mean net worth of ($1,400)
� median net worth of $1,700.
� 10% of HH had financial assets of <$100
� In terms of net financial assets, 25 to 33%are asset poor
� can’t survive three months at poverty line
Source: Bucks, Kennickell, and Moore (2006), SCF (2004), Havenmann and Wolff (2001), Caner and Wolf (2002)
© 2007, D2D Fund, Inc. 15
Estimates of U.S. Asset Poverty by
Group
� Blacks / Hispanics 62 percent
� Head aged less than 25 years 72 percent
� Head aged 25-34 years 52 percent
� Head with < high school degree 60 percent
� Renters 64 percent
� Female heads with children 71 percent
Source: Havenmann and Wolff (2001).
© 2007, D2D Fund, Inc. 16
Why Target LMI Savings?
� Need (as above)
� Impact
� Same $$$ / higher marginal impact on LMI families
� Assets matter - to individuals, families, kids,communities (Sherraden 1991, 2001)
� Appetite
� While savings rates are down, not so for lowest 60%of Americans (Maki and Palumbo 2001)
� When given incentives to save, poorest familiessaved at highest rates (Sherraden 2007)
© 2007, D2D Fund, Inc. 17
Efforts & Barriers to Boost LMI Saving
� “Asset Building” field has emerged� champion of LMI saving / asset building
� laboratory for policies, programs, products
� But scale, access & cost are still problems
� Innovation just beginning – mixed success
� Products: SVCs, HRB “Easy IRAs”
� Marketing: America Saves (social marketing), ZMET
� Policy: split refunds, savers’ credit, Automatic IRAs
© 2007, D2D Fund, Inc. 18
Saving Challenges for LMI
Uneasy
About
Saving
Not
Allowed
to Save
Not
Thinking
About
Saving
� Competingspendingpriorities
� Need accessto $ (liquidity)
� ChexSystems
� Minimumbalance req.
� No local FIpresence
� Mistrustdepositories
� Risk averse
� Skeptical
� Unclear goal
© 2007, D2D Fund, Inc. 19
Product Challenges Especially Tough
Barrier
� Small balances
� Checkered (credit) past
� Risk averse
� Mistrust mainstreamfinancial service firms
� No bank account
� Unclear goal(s)
� Long-term goals &liquidity needs in tension
Bond Feature
� Low $50 minimum
� No credit screen
� Principal is risk free
� Gov’t backed, no fees,good return
� No bank acct required
� Can save for anything
� Long-term orientation,access after 1 year
© 2007, D2D Fund, Inc. 20
Other Key Bond Features
� Gift-able
� can be purchased for self or with co-owner
� Inflation indexed
� Strong history
� high awareness
� fond memories
� No Patriot Act or KYC compliance
© 2007, D2D Fund, Inc. 21
How Savings Bonds Compare
YesNoNoMay Buy as a Gift
NoYesYesChexSystems Review
3 months interest(redemptions within 5years of purchase)
Greater of 3 monthsinterest or $25
NoneEarly RedemptionPenalty / Forfeiture
1 yearNoneNoneMinimum Holding Period
Adjusts bi-annuallyYesNoRate Fixed
$50$2,500$100Minimum to Open
$0$0$36Annual Fees
4.52%3.25%0.20%Rate of Return
Series I U.S.Savings Bond
Example 1 YearCertificate of Deposit 2
Example SavingsAccount1
1) Source Bank of America (www.bofa.com), Regular Savings Account as of 6/15/07;annual fees assumes balance less than $300.
2) Source: Wells Fargo (www.wellsfargo.com); Standard CD as of 6/15/07; rate ofreturn assumes balance less than $5,000.
© 2007, D2D Fund, Inc. 22
Savings Bond History
� Today’s savings bonds trace roots to 1935
� for small saver especially� “appeal primarily to individuals w/ small amounts to invest”
(U.S. Department of the Treasury, 1981)
� for public good, not just gov’t borrowing needs� Dept. of Treasury “to promote wise spending, intelligent
saving, & safe investment” (U.S. Department of theTreasury, 1918)
� to “democratize public finance”� “Every (person) who owned a Gov’t Bond... would serve
as a bulwark against the constant threats to Uncle Sam’spocketbook from pressure blocs & special-interest groups”(Morgenthau, 1944)
© 2007, D2D Fund, Inc. 23
Savings Bond Marketing
© 2007, D2D Fund, Inc. 24
Fundamental Questions:
Demand? New Savings?
� Do LMI individuals want savings bonds?
� If so, who, why & under what circumstances?
� Would bond sales constitute new or substitutesaving?
� Is there demand, does it increase savings
� One way to find out: experiment
© 2007, D2D Fund, Inc. 25
TS 06: Pre-experiment
� 2006 pre-experiment� location - 1 HRB district, 14 offices (Schaumberg, IL)� duration - 3 weeks only� context - commercial tax preparer only� purpose - test operations, indication of demand
� Take-aways:
� 3% take up rate despite handicaps
� no gifting option
� late in tax season
� limited training
� operationally complex - need to streamline
© 2007, D2D Fund, Inc. 26
TS 2007 Savings Bond Experiment
� Product: Series I bonds� fixed rate (premium above inflation) + inflation rate
� fixed rate set at purchase, inflation rate adjustsevery 6 months
� Channels� H&R Block offices
� larger scale, commercial firm environment
� select VITA (Volunteer) tax prep. sites� lower income clients, core mission to support saving
� Dates� Tax season start till March (VITA), April (HRB)
© 2007, D2D Fund, Inc. 27
Pilot Details: Marketing & Sales Training
H&R Block
� Relied on taxpreparers (“pros”) tooffer bonds to clients
� Majority of 400 taxpros trained in groupsessions
� Marketing limited topoint of sale (posters,monitor toppers, etc.)
VITA
� “Bond Coordinator”marketed to clients,dealt w/ ops issues
� Most tax preparerstrained directly orindirectly (used CBT)
� Marketing limited topoint of sale (posters,brochures)
© 2007, D2D Fund, Inc. 28
Marketing Examples
VITA H&R Block
© 2007, D2D Fund, Inc. 29
Research Tools
H&R Block
� Tax data
� Survey data
� buyers
� non-buyers
� control
� Transaction data
� savings & settlementproducts
VITA
� Tax data
� Survey data
� buyers
� sample of non-buyers
� Transaction data
� bond orders details
© 2007, D2D Fund, Inc. 30
Research Details
� Consents from all parties
� All personal identifiers stripped from taxdata
� Form 7216 authorization obtained fromall H&R Block clients
� Research participation not required forbond purchase
© 2007, D2D Fund, Inc. 31
Bond Eligibility - Requirements to Buy
H&R Block
� Refund >$500
� Willing to open EasySavings account tofacilitate buying bond
� total bond sale > $50
VITA
� Refund >$50
� Willing / able to usedirect deposit (orWestern Union QuickCash product)
� total bond sale > $50
© 2007, D2D Fund, Inc. 32
Bond Purchase Processing
H&R Block
� Individual “conduits”(Easy Savings Acct)
� HRB splitting(internal systems)
� Instructions, consentcollected via software
� Difficult to train taxpros & explain toclients, automatedprocessing
VITA
� Pooled “conduit”(clearing account)
� IRS splitting (via newForm 8888)
� Instructions, consentcollected via paper;faxed by site
� Easy to train preparers,labor-intensive, errorprone processing
© 2007, D2D Fund, Inc. 33
What Did We Find?
� Context
� roughly 10,000 study participants
� Block: 3,729 Treatment / 1,485 Control
� VITA: 4,410 Treatment / 431 “Incentive Offer”
� Block
� 27 offer & 4 control offices
� 2 cities (Boston, Schaumberg, IL)
� VITA
� 4 offices
� 4 cities (Baltimore, Boston, Denver, Tulsa)
© 2007, D2D Fund, Inc. 34
Measuring Take Up
� Take up = purchase / offer
� Nuances in determining eligible offers
� Block
� definitive offers - certain client offered bond
� possible offers - possible client offered bond
� VITA
� direct depositers - clients who had sufficient refund($50) & used direct deposit (required to buy bonds)
� all eligible - clients who had sufficient refund ($50)to buy a bond
© 2007, D2D Fund, Inc. 35
Conditional on receiving an offer to buy bonds:
Does Anyone Want Bonds?
Block
� 5.9% bought (definitive)
� 1.1% bought (possible)
� 220 buyers
� 407 bond recipients
VITA
� 9.6% (direct
depositers)
� 6.0% (all eligible)
� 231 buyers
� 377 bond recipients
© 2007, D2D Fund, Inc. 36
Compared to What?
� For H&R Block offices:
IRA
Openings & Re-contributions
Savings Account Openings & Re-
contributions
Treatment offices 1.5% 1.4%
Control offices 0.6% 0.1%
Block national 0.9% 0.2%
© 2007, D2D Fund, Inc. 37
Compared to What? (Other Samples)
� Opened IRAs (w/out incentive) at tax time
� H&R Block experiment, 2005 (Duflo et al.)
� 2.2% (all filers, pre-experiment)
� 3.5% (non-married, avg. AGI 30k)
� 2.9% (married filers, avg. AGI 69k)
� Contributed to an IRA� Percent of eligible tax filers (IRS data, TY02)
� 3.4% (AGI < 20k)
� 4.6% (AGI < 30k)
� 5.6% (AGI < 40k)
© 2007, D2D Fund, Inc. 38
Explaining Demand
� Bonds are familiar to buyers� Awareness: 85-88% (Block), 73-75% (VITA)
� Prior purchase: 35-39% (Block), 15% (VITA)
Have you ever purchased a US Savings Bond for yourself and/or
someone else before?
0 %
10 %
2 0 %
3 0 %
4 0 %
5 0 %
6 0 %
7 0 %
A . Y e s , f o r
my s e lf .
B . Y e s , f o r
s o me o ne e ls e
C . Y e s , f o r
my s e lf A N D
s o me o ne e ls e
D . N o , I ha v e
ne v e r p urc ha s e d a
b o nd .
someone else before?
3 0 %
4 0 %
50 %
6 0 %
70 %
Bond Purchasers
Non-BondPurchasers
© 2007, D2D Fund, Inc. 39
Take-up in Context
� Long-term product� Bonds connote long-term saving
� Requirement for 1 year holding, minimum
� No immediate tax benefit (vs. IRA)� Nothing for tax preparer to “sell” to client
� Took patience, endurance to buy a bond� Block
� open another product (Easy Savings)� get balance paid back to you (orders under $300)
� VITA� complete 3 additional forms� entrust part of refund to bank not present at site
© 2007, D2D Fund, Inc. 40
What if We Offer an Incentive?
� 20% discount on bond purchase
� for 1 week (1/29 - 2/3) in Tulsa VITA site
� goal: gauge effect of financial incentive
� up to $70 subsidy ($350 purchase) per client
� special marketing materials highlightedincentive
� 431 clients offered bonds w/ incentive
� 60 takers
© 2007, D2D Fund, Inc. 41
Incentive Findings
� Take-up: rate nearly doubled or more� 15.2% of those offered & eligible
� 18.1% of direct depositers
� Same portion of refund spent (5.8%) as non-incentive;avg $ / client $189 vs. $185
� Only 22% “yes, would’ve bought bonds w/o incentive”
� Compare to IRAs opened at tax time� H&R Block experiment, 2005 (Duflo et al.)
� 20% incentive (added to client contribution)
� 9.6% take up of eligible tax filers ($300 refund)
� Average AGI $44,789 (IRA) vs. $20,704 (bond)
© 2007, D2D Fund, Inc. 42
Block VITA (averages)
Buyers Non-Buyers Buyers Non-Buyers
n 220 3,509 231 4,179
AGI $ 36,828 $ 41,171 $ 20,888 $ 18,105
Refund $ 3,836 $ 2,951 $ 3,177 $ 1,373
Dependents 1.3 0.8 1.3 0.6
% of Refund Used for Bonds
5.9% - 5.8% -
% of Bonds w/ Co-Owners
82% - 84% -
Buyer / Non-Buyer Details
� Huge portion w/ co-owners� Buyers appear to have
� more dependents� larger refunds
� Consistent % of refund used for bonds
© 2007, D2D Fund, Inc. 43
Gifting: Co-ownership
� Bonds may be purchased with:� beneficiaries (contingent on owner’s death)
� co-ownership (either party may redeem bond)
� Chose co-ownership as most gift-like� only name required (no SSN) - easy point of sale
� up to 4 (Block) / 5 (VITA) co-owners (plus client)
� buyers: a) self only, b) self & others, c) others only
© 2007, D2D Fund, Inc. 44
Gifting Details
� Among VITA bond buyers:� 59% bought one or more bonds for child
� 9% bought one or more for grandkids
� Among Block bond buyers:
� Most popular to buy bonds as gift by wide margin
� “Self & Others” spent the most on bonds, boughtmost gifts
Bought for
SELF only OTHERS only SELF & OTHERS
% of purchasers 31% 52% 17%
Avg $ spent / buyer $ 145 $ 241 $ 335
Avg. # of gifts / buyer - 1.9 2.3
© 2007, D2D Fund, Inc. 45
Gifting: What’s Going on?
� D2D consumer research suggests childrenare a powerful force in LMI financial plans
© 2007, D2D Fund, Inc. 46
More on Gifting - Gender Differences
© 2007, D2D Fund, Inc. 47
New Savings?
� For bond buyers:� 54% (VITA) / 40% (Block)
reported no prior savings
� Of remainder, 46% (VITA) / 22%(Block) reported savings &investments < $1,000
� 53% (VITA) said they wouldhave saved less if they had notbought a bond
No
40%Yes
56%
4 %
Block Bond Buyers:
Do you have any savings
or investments?
© 2007, D2D Fund, Inc. 48
Why Did This Work? The Good News
� Strong foundation� very high awareness, experience w/ bonds
� solid product design (return, fees, credible)
� Bond features appeal to LMI tax clients� gifting
� small amounts (1/3 of VITA bonds @ $50)
� allows small savings “win”?
� Tax time & bonds a good fit� immediacy & simplicity (do it now / it’ easy)
� compliance relatively easy (Pat Act, KYC)
© 2007, D2D Fund, Inc. 49
Bank #1
Bank #2
Tax Filer
Bank #3
-- Bank
Midwest
IRS
Bureau of
Public Debt
Tax Pro enters
coupon into TPS for bond
type purchase
H&R Block
Tax Pro fills out
purchase form
Signs new
account agree-
ment and consent
Tax Pro
faxes form to
WHQ for process-
ing Client info
added to refund card
account
WHQ monitors
card account
for refund
receipt
WHQ sends
letter to client
outlining purchase details
WHQ fills out
spread-sheet with
purchase info and
uploads
Move funds from
pooled
account to fed
account
Submit via
norma l savings
bond process
Contact HRB to indicate bonds
purchase complete
Tax prep and
return filing
Receive Savings Bonds
info (via
Fed Line)
Process Savings
Bond
Mail Bond to Tax Filer
Receive US
Savings
Bond
Process Tax
Return
Send Refund
Send Instruct-
ions for Temp-orary
Account Opening
Send Instruct-
ions for Split
between debit and
check
Open Temp-orary
Account
Split the Refund
Send Funds to
HRB refund card
account
Send Check to Tax Filer
Close Account
Receive Refund
Check
Decide to buy
bond(s)
Funds received & moved (via ach)
to a pooled
account
Refund card
account closed
Receive bond
purchase
letter
Tax Season 2006, US Savings Bond Process Flow
Bank #1
Bank #2
Tax Filer
Bank #3
-- Bank
Midwest
IRS
Bureau of
Public Debt
Tax Pro enters
coupon into TPS for bond
type purchase
H&R Block
Tax Pro fills out
purchase form
Signs new
account agree-
ment and consent
Tax Pro
faxes form to
WHQ for process-
ing Client info
added to refund card
account
WHQ monitors
card account
for refund
receipt
WHQ sends
letter to client
outlining purchase details
WHQ fills out
spread-sheet with
purchase info and
uploads
Move funds from
pooled
account to fed
account
Submit via
norma l savings
bond process
Contact HRB to indicate bonds
purchase complete
Tax prep and
return filing
Receive Savings Bonds
info (via
Fed Line)
Process Savings
Bond
Mail Bond to Tax Filer
Receive US
Savings
Bond
Process Tax
Return
Send Refund
Send Instruct-
ions for Temp-orary
Account Opening
Send Instruct-
ions for Split
between debit and
check
Open Temp-orary
Account
Split the Refund
Send Funds to
HRB refund card
account
Send Check to Tax Filer
Close Account
Receive Refund
Check
Decide to buy
bond(s)
Funds received & moved (via ach)
to a pooled
account
Refund card
account closed
Receive bond
purchase
letter
Tax Season 2006, US Savings Bond Process Flow
What Worked? The Bad News
� Operational nightmare - it took thisto allow bond sales in commercialtax preparer
© 2007, D2D Fund, Inc. 50
Aren’t All Essential Ingredients
Available?
++++
Savings Bonds + IRS Split Refunds
Why isn’t this enough?
© 2007, D2D Fund, Inc. 51
Why Not Enough?
� Someone has to ask� what would take up look like if no one asked
� solution: make it so all tax filers likely to consider it
� Firms won’t process at scale� complex = expense
� solution: facilitate process so private sector doesn’thave to subsidize government when consumerasks the government to “Just keep my money”
© 2007, D2D Fund, Inc. 52
Not a New Idea: Bonds on Tax Forms
1962 IRS
Form
1040A
© 2007, D2D Fund, Inc. 53
What Can We Do?
� Continue to facilitate gifting� any changes must allow (promote?) gifting
� Ensure bonds do no harm� don’t discourage saving in bonds by including bonds
in public assistance means testing
� Align bonds w/ realities of LMI households� expand & make emergency policy more explicit
� recognize financial lifecycle (rollovers)
� Re-brand around families (“family bonds”)
� Consider financial incentives� for LMI to buy bonds
� for tax preparers to claim on clients’ behalves
© 2007, D2D Fund, Inc. 54
Policy Agenda Framework: 3 Pillars
� Increase & facilitate access, appeal� offer in tax filing process, with gifting
� clarify / expand emergency redemption policy
� exclusion for means tested benefits
� Maintain / Build Awareness� restore marketing budget
� re-brand around families, gifting
� Offer Incentives� expand Saver’s Credit
� Series T Bonds - at a discount, at tax time only, forEITC eligible
© 2007, D2D Fund, Inc. 55
Contact
Prof. Peter Tufano
Harvard Business School / D2D Fund, Inc.
617.495.6855
Timothy Flacke
D2D Fund, Inc.
www.d2dfund.org
617.541.9064