Download:
pdf |
pdfF E D E R A L D E P O S I T I N S U R A N C E C O R P O R AT I O N
2017
FDIC National Survey of
Unbanked and Underbanked
Households
ECONOMICINCLUSION.GOV
2017
FDIC National Survey of
Unbanked and Underbanked
Households
OCTOBER 2018
Members of the FDIC Unbanked/
Underbanked Survey Study Group
Gerald Apaam
Susan Burhouse
Karyen Chu
Keith Ernst
Kathryn Fritzdixon
Ryan Goodstein
Alicia Lloro
Charles Opoku
Yazmin Osaki
Dhruv Sharma
Jeffrey Weinstein
FEDERAL DEPOSIT INSURANCE CORPORATION
Division of Depositor and Consumer Protection
ECONOMICINCLUSION.GOV
2017 FDIC National Survey of Unbanked and Underbanked
Households
Table of Contents
1.
Executive Summary................................................................................................................................................. 1
2.
Background and Objectives.................................................................................................................................. 15
3.
Banking Status of U.S. Households...................................................................................................................... 17
4.
Banked Households: Types of Accounts, Methods Used to Access Accounts, and Bank Branch Visits............. 25
5.
Prepaid Cards........................................................................................................................................................ 34
6.
Alternative Financial Services................................................................................................................................ 39
7.
Saving for Unexpected Expenses or Emergencies................................................................................................ 43
8.
Credit..................................................................................................................................................................... 48
9.
How Households Conduct Their Financial Transactions in a Typical Month......................................................... 56
10. Measuring Economic Inclusion.............................................................................................................................. 61
11. Implications and Conclusions................................................................................................................................ 67
Appendix 1. FDIC Technical Notes................................................................................................................................. 71
Appendix 2. 2017 Revisions to the FDIC National Survey of Unbanked and Underbanked Households...................... 75
Appendix 3. Survey Instrument...................................................................................................................................... 77
Appendix Tables A – H (published separately)
V
2017 FDIC National Survey of Unbanked and Underbanked
Households
1. Executive Summary
The FDIC is committed to expanding Americans’ access
to safe, secure, and affordable banking services. The FDIC
National Survey of Unbanked and Underbanked Households
is one contribution to this end.
To assess the inclusiveness of the banking system, and in
partial response to a statutory mandate, the FDIC has conducted the survey biennially since 2009.1 The most recent
survey was administered in June 2017 in partnership with the
U.S. Census Bureau, collecting responses from more than
35,000 households. The survey provides estimates of the
proportion of U.S. households that do not have an account at
an insured institution and the proportion that have an account
but obtained (nonbank) alternative financial services in the
past 12 months. The survey also provides insights that may
inform efforts to better meet the needs of these consumers
within the banking system.
This executive summary presents key results from the 2017
survey and summarizes the implications of these results for
policymakers, financial institutions, and other stakeholders
who are working to improve access to mainstream financial
services.
Banking Status of U.S. Households
• In 2017, 6.5 percent of U.S. households were “unbanked,”
meaning that no one in the household had a checking or
savings account. The unbanked rate in 2017 declined to
the lowest level since the survey began in 2009. Since the
survey was last administered in 2015, the unbanked rate
has fallen by 0.5 percentage points.
»» A
pproximately 8.4 million U.S. households, made up
of 14.1 million adults and 6.4 million children, were
unbanked in 2017.2
Figure ES.1 National Estimates, Household Unbanked
Rates by Year
7.6
2009
8.2
2011
7.7
2013
7.0
2015
6.5
2017
• An additional 18.7 percent of U.S. households were “underbanked” in 2017, meaning that the household had an
account at an insured institution but also obtained financial products or services outside of the banking system.
Specifically, a household is categorized as underbanked if
it had a checking or savings account and used one of the
following products or services from an alternative financial services (AFS) provider in the past 12 months: money
orders, check cashing, international remittances, payday
loans, refund anticipation loans, rent-to-own services,
pawn shop loans, or auto title loans.
»» A
pproximately 24.2 million U.S. households, composed
of 48.9 million adults and 15.4 million children, were
underbanked in 2017.
»» The underbanked rate in 2017 was 1.2 percentage
points lower than the 2015 estimate (19.9 percent).
• Almost 70 percent (68.4 percent) of U.S. households were
“fully banked” in 2017, meaning that the household had a
bank account and did not use AFS in the past 12 months.
The fully banked rate in 2017 was slightly higher than the
2015 estimate (68.0 percent).
Section 7 of the Federal Deposit Insurance Reform Conforming Amendments Act of 2005 (Pub. L. 109 – 173) calls for the FDIC to conduct ongoing surveys, “on
efforts by insured depository institutions to bring those individuals and families who have rarely, if ever, held a checking account, a savings account or other type of
transaction or check cashing account at an insured depository institution [‘unbanked’] into the conventional finance system.” Section 7 further instructs the FDIC to
consider several factors when conducting the surveys, including estimating the size and worth of the unbanked market in the United States and identifying the primary
issues that prevent unbanked individuals from establishing conventional accounts.
1
Adults are defined as people aged 16 and older. The estimates of 14.1 million adults and 6.4 million children may understate the total number of people in the United
States who do not have access to a bank account because these figures do not include residents of “banked” households who do not have an account in their name
and do not benefit from a bank account owned by another household resident.
2
1
Table ES.1 National Estimates, Household Banking Status by Year
For all households, row percent
Year
Number of Households
(1000s)
Unbanked
(Percent)
Underbanked
(Percent)
Fully banked
(Percent)
Banked, underbanked
status unknown
(Percent)
2013
123,750
7.7
20.0
67.0
5.3
2015
127,538
7.0
19.9
68.0
5.0
2017
129,276
6.5
18.7
68.4
6.3
Changes in Banking Status
• T
he decline in the unbanked rate from 2015 to 2017 can
be explained almost entirely by changes in household
characteristics across survey years, particularly improvements in the socioeconomic circumstances of U.S. households. After accounting for these changes, the remaining
difference in the unbanked rate from 2015 to 2017 was
very close to zero and no longer statistically significant.3
• Consistent with previous surveys, banking status in 2017
varied considerably across the U.S. population. For example,
unbanked and underbanked rates were higher among lower-income households, less-educated households, younger
households, black and Hispanic households, working-age
disabled households, and households with volatile income.4
• Unbanked rates in 2017 were lower than or similar to
unbanked rates in recent years for most segments of the
population.
»» R
ecent declines in unbanked rates have been particularly sharp for younger households, black households,
and Hispanic households.5 Despite these improvements,
unbanked rates for these groups remained substantially
higher than the overall unbanked rate in 2017.
»» Unbanked rates did not decline in recent years for a
few segments of the population. For example, among
working-age disabled households, unbanked rates
were similar in 2013, 2015, and 2017.
• Reflecting the decline in the underbanked rate at the national level between 2015 and 2017, underbanked rates
also declined for many segments of the population during
that period.
»» For example, underbanked rates decreased for
households with less than $15,000 in income,
households with a high school diploma (but no
college), and working-age disabled households.
Unbanked Households: Previous Banking Status
and Future Banking Plans
s discussed in previous reports, bank account ownership is
A
not static and some households appear to cycle in and out of
the banking system.
• Nearly
half of unbanked households in 2017 had a bank
account at some point in the past, similar to previous years.
• The proportion of unbanked households that were “very
likely” or “somewhat likely” to open an account in the next
12 months declined in 2017 compared with earlier years,
while the proportion that were “not at all likely” increased.
»» One in four unbanked households in 2017 were very
likely or somewhat likely to open an account, down
from 37.9 percent in 2013.6
»» More than half (58.7 percent) of unbanked households in
2017 were not at all likely to open an account, up from
40.0 percent in 2013. This increase was fairly widespread
among segments of the unbanked population.7
• As in previous years, interest in opening an account in the
next 12 months was higher among unbanked households
that had a bank account at some point in the past, compared
with unbanked households that never had an account.
A linear probability model was estimated to account for changes from 2015 to 2017 in the distribution of households across the household characteristics listed in
Appendix Table A.2. Changes in the socioeconomic characteristics of households (annual income level, monthly income volatility, employment status, homeownership
status, and educational attainment) between 2015 and 2017 accounted for almost all of the difference in unbanked rates between 2015 and 2017. Adding controls for
the remaining demographic characteristics listed in Appendix Table A.2 had little effect on the remaining difference.
3
For characteristics that vary at the person-level, such as race, age, and education, the characteristics of the owner or renter of the home (i.e., householder) are used to
represent the household. For convenience, abbreviated language is used when referring to certain household characteristics. For example, the term “white household”
refers to a household in which the householder has been identified as white, non-black, non-Hispanic, and non-Asian. The phrase “working-age disabled household”
refers to a household in which the householder has a disability and is aged 25 to 64. See Appendix 1 for additional details. For monthly income volatility, the 2015 and
2017 surveys asked households whether their income over the past 12 months “was about the same each month,” “varied somewhat from month to month,” or “varied
a lot from month to month.” The term “volatile income” refers to a household with income that varied somewhat or a lot from month to month.
4
The decline in the unbanked rate for black households from 2013 to 2017 was no longer statistically significant after accounting for changes in the other household
characteristics listed in Appendix Table A.2 (except for monthly income volatility, which is not available for 2013). Most of the decline can be attributed to changes in
income and the other household characteristics across survey years.
5
Estimates of the likelihood of opening a bank account in the next 12 months for 2013 and 2015 differ from those published in earlier reports because observations
with missing information on the likelihood of opening a bank account in the next 12 months were not dropped in earlier reports.
6
The proportion of unbanked households that were not at all likely to open an account in the next 12 months was substantially higher in 2017 than in 2013, even after
accounting for changes in the household characteristics listed in Appendix Table A.2 (except for monthly income volatility, which is not available for 2013) and in the
use of prepaid cards between 2013 and 2017.
7
2 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Figure ES.2 Unbanked Rates by Household Age and Year
15.7
13.1
12.5
10.0
10.6
8.5
9.0
8.9
7.8
7.5
6.7
6.9
5.6
5.8
5.9
3.5
15 to 24 years
25 to 34 years
35 to 44 years
2013
45 to 54 years
2015
55 to 64 years
3.9
3.1
65 years or more
2017
Figure ES.3 Unbanked Rates by Household Race and Ethnicity and Year
20.6
18.2
16.9
17.9
16.2
15.0
14.0
12.8
11.1
4.0
Black
Hispanic
3.6
2.5
2.2
Asian
2013
3.1
3.0
White
2015
Other
2017
Table ES.2 Unbanked Households’ Likelihood of Opening a Bank Account in Next 12 Months by Year
For all unbanked households, row percent
Year
Number of
Unbanked
Households
(1000s)
Very likely
(Percent)
Somewhat likely
(Percent)
Not very likely
(Percent)
Not at all likely
(Percent)
2013
9,021
14.6
23.3
22.1
40.0
2015
8,358
10.2
18.2
19.4
52.2
2017
7,682
9.5
15.6
16.3
58.7
3
Reasons Households Were Unbanked
As in previous years, the 2017 survey asked unbanked
households about the reasons why they did not have a bank
account. Findings are similar to those reported in previous
years.
• M
ore than half (52.7 percent) of unbanked households
cited “Do not have enough money to keep in an account”
as a reason for not having an account, the most commonly
cited reason. This reason was also the most commonly
cited main reason for not having an account (34.0 percent).
• Almost one-third (30.2 percent) of unbanked households
cited “Don’t trust banks” as a reason for not having an
account, the second-most commonly cited reason. This
reason was also the second-most commonly cited main
reason (12.6 percent).
• Higher proportions of unbanked households that were not
at all likely or not very likely to open a bank account in the
next 12 months cited “Don’t trust banks” (36.2 and 31.5
percent, respectively) in 2017, compared with unbanked
households that were somewhat likely or very likely to
open a bank account in the next 12 months (24.7 and 21.0
percent, respectively).
Types of Accounts Owned by Banked Households
• S
avings and checking account ownership among banked
households in 2017 was similar to previous years.
»» Almost all banked households had a checking account
(98.2 percent), while roughly three in four (78.0 percent)
had a savings account.
»» Savings account ownership rates in 2017 varied widely
across the population. For example, savings account
ownership rates were lower among lower-income
households, less-educated households, Hispanic
households, working-age disabled households, and
households in rural areas.
• As in previous years, higher proportions of unbanked
households that previously had an account cited “Bank
account fees are too high” (29.9 percent) or “Bank
account fees are unpredictable” (24.9 percent) in 2017,
compared with unbanked households that never had an
account (21.1 and 17.0 percent, respectively).
Figure ES.4 Reasons for Not Having a Bank Account, Unbanked Households, 2017 (Percent)
Do not have enough money to keep in account
34.0
Don't trust banks
Avoiding bank gives more privacy
Inconvenient hours
Inconvenient locations
Other reason
24.7
8.6
20.2
1.3
ID, credit, or former bank account problems
Banks do not offer needed products or services
28.2
3.0
Account fees too high
Account fees unpredictable
30.2
12.6
14.0
4.9
13.1
1.6
4.1
2.1
9.4
9.2
14.9
12.3
Cited
4 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Main
52.7
Methods Banked Households Used to Access Their
Accounts
Bank Branch Visits Among Banked Households
The 2017 survey included new questions that asked all
households whether they spoke with a teller or other employee in person at a bank branch (i.e., visited a bank branch)
in the past 12 months, and if so, how many times. Since
2013, the survey has measured the share of households that
accessed their account using bank tellers. However, some
households may rely on bank branches for activities other
than accessing an account, such as resolving a problem or
asking about products or services, and the questions on account access methods provide only an imprecise measure of
the intensity of branch use. The goal of the new questions is
to provide a more complete picture of household use of bank
branches.
Use of mobile banking to access a bank account continued
to increase sharply, while use of bank tellers declined. Use of
bank tellers, however, remained quite prevalent, particularly
among segments of the population that had higher unbanked
and underbanked rates.
• The proportion of banked households that used mobile
banking to access their accounts in the past 12 months increased from 23.2 percent in 2013 to 31.9 percent in 2015
and 40.4 percent in 2017. The share of banked households that used mobile banking as their primary method of
account access also increased sharply from 2013 to 2017,
both overall and across household characteristics.
• Overall, 86.0 percent of banked households visited a bank
branch in the past 12 months, and 35.4 percent visited ten
or more times.8
• In 2017, almost three in four (73.6 percent) banked households used bank tellers to access their accounts in the
past 12 months, a higher proportion than any other method asked about in the survey. However, use of bank tellers
declined modestly between 2013 and 2017. The share of
banked households that used bank tellers as their primary
method of account access decreased substantially, both
overall and across household characteristics; however,
this method is still the second-most prevalent primary
method overall after online banking.
• Branch visits were prevalent even among banked households that used online or mobile banking as their primary
method of account access. For example, 81.0 percent of
banked households that used mobile banking as their primary method visited a branch in the past 12 months, and
nearly one-quarter (23.0 percent) visited ten or more times.
• Patterns of bank branch visits among banked households
varied by household characteristics. For example, older
households, households in rural areas, and households
with volatile income were more likely to visit a branch or to
have visited ten or more times. Black, Hispanic, and Asian
households were less likely to visit a branch or to have
visited ten or more times.
»» Use of bank tellers as the primary means of account
access remained quite prevalent among certain
segments of the population, including lower-income
households, less-educated households, older households, and households in rural areas. These groups
were also disproportionately more likely to access their
accounts using only bank tellers.
Table ES.3 All Methods Used to Access Bank Accounts by Year
For all banked households that accessed their account in the past 12 months, row percent
Year
Number of
Households
(1000s)
Bank teller
(Percent)
ATM/Kiosk
(Percent)
Telephone
banking
(Percent)
Online
banking
(Percent)
Mobile
banking
(Percent)
Other
(Percent)
2013
108,295
78.8
69.6
26.1
55.1
23.2
1.0
2015
113,315
75.5
69.8
27.0
60.4
31.9
1.1
2017
115,040
73.6
71.6
28.9
63.0
40.4
0.9
Note: Row percentages sum to more than 100 because households were asked to select all bank account access methods used.
Table ES.4 Primary Method Used to Access Bank Accounts by Year
For all banked households that accessed their account in the past 12 months, row percent
Year
Number of
Households
(1000s)
Bank teller
(Percent)
ATM/Kiosk
(Percent)
Telephone
banking
(Percent)
Online
banking
(Percent)
Mobile
banking
(Percent)
Other
(Percent)
2013
108,295
32.2
24.4
3.3
32.9
5.7
0.8
2015
113,315
28.2
21.0
3.0
36.9
9.5
0.9
2017
115,040
24.3
19.9
2.9
36.0
15.6
0.7
Among unbanked households, 14.7 percent visited a bank branch in the past 12 months: 7.7 percent visited a branch one to four times, 2.2 percent visited five to nine
times, and 4.7 percent visited ten or more times. Approximately two-thirds of unbanked households that visited a branch did not have a bank account at any time in
the past 12 months.
8
5
Figure ES.5 Bank Branch Visits in Past 12 Months Among
Banked Households, 2017 (Percent)
35.4
30.8
»» Other common activities, performed by approximately
one-third of banked households in 2017, were using a
bank’s mobile website or bank’s mobile app to check
a bank account balance or recent transactions, and
receiving a mobile text alert or push notification from a
bank about an account.
18.2
14.0
0 times
• Use of a mobile phone to check email from a bank about an
account was the most common activity in 2017, performed
by 44.1 percent of banked households.
1 to 4
times
5 to 9
times
10 or more
times
»» The remaining mobile activities asked about in the survey were less common, but the proportion of banked
households that performed each of these activities
doubled or more than doubled from 2013 to 2017.
Note: Households that visited a branch but with unknown frequency
(1.6 percent of banked households) are not shown.
Mobile Activities Among Banked Households
The 2017 survey included a series of questions about the
ways households used a mobile phone for banking activities
in the past 12 months. Most of these activities were also
asked about in the 2013 survey.
• Underbanked households were more likely to perform
each mobile activity than fully banked households. Use
of each mobile activity was also more common among
higher-income households, more-educated households,
younger households, working-age nondisabled households, and households with volatile income.
Figure ES.6 Mobile Activities Among Banked Households by Year (Percent)
Checked email about an account
44.1
19.0
Checked balance or transactions
35.4
Text message alert
34.0
13.2
Bill payment
26.5
12.2
Transferred money between accounts
Deposited a check electronically
Sent money to others
25.4
5.6
18.0
5.9
13.7
2013
2017
Note: Estimates of the proportion of banked households that used a mobile phone to check email from a bank about an account or that received a mobile text alert or
push notification from a bank about an account are not available for 2013.
6 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Prepaid Cards
Figure ES.7 Prepaid Card Use in Past 12 Months by
Banking Status and Year (Percent)
ome consumers use general purpose reloadable prepaid
S
cards to address their financial transactions needs. Similar
to a checking account, these cards can be used to pay bills,
withdraw cash at ATMs, make purchases, deposit checks,
and receive direct deposits. Consumers can obtain prepaid
cards from sources such as a bank location or bank’s website, a nonbank store or website, a government agency, or an
employer. Many, although not all, such cards store funds in
accounts eligible for deposit insurance.9
27.1 26.9
22.3
13.1
15.4 14.5
5.3
Unbanked
Underbanked
2013
• Between 2015 and 2017, the proportion of households
that used prepaid cards decreased from 9.8 percent to 9.2
percent. This decline can be attributed primarily to changes in income and other characteristics of U.S. households
between 2015 and 2017. However, the proportion of
households that used prepaid cards in 2017 remained
higher than in 2013 (7.9 percent).
2015
6.9
6.7
Fully banked
2017
»» Unbanked households that used prepaid cards were
more likely to have had a bank account at some point
in the past: 62.7 percent of unbanked households that
used prepaid cards in 2017 had a bank account in the
past, compared with 41.9 percent of unbanked households that did not use prepaid cards.
• As in previous years, prepaid card use in 2017 was higher
among lower-income households, less-educated households, younger households, black households, working-age disabled households, and households with volatile
income.
• Consistent with previous survey results, households that
used prepaid cards in 2017 obtained them from a variety
of sources. The most common source in 2017 was a store
or website that is not a bank, followed by a government
agency, family or friends, and a bank location or a bank’s
website.
• Use of prepaid cards in 2017 was most prevalent among
unbanked households, as in previous years.
Figure ES.8 Sources of Prepaid Cards for Households That Used Prepaid Cards in Past 12 Months by Year (Percent)
42.6
Store or website that is not a bank
45.4
Government agency
14.8
15.0
Family or friends
14.2
15.0
17.3
Bank location or bank's website
13.3
9.2
9.3
Employer payroll card
6.8
8.4
Other
Unknown
1.3
0.6
2015
2017
Note: Bars sum to more than 100 percent because households with multiple prepaid cards were asked to select all sources of their cards.
Unless noted otherwise, estimates of prepaid card use are based on the 12 months before the survey. Households were instructed that the survey questions about
prepaid cards were “not asking about gift cards or debit cards linked to a checking account.”
9
7
Alternative Financial Services
Saving for Unexpected Expenses or Emergencies
• In 2017, 22.1 percent of households used some type of
AFS in the past 12 months, down from 24.0 percent in
2015 and 24.9 percent in 2013.10
Savings can help households better manage unexpected expenses or emergencies, such as a sudden illness, job loss, or
home or car repairs. The absence of savings can sometimes
be a barrier to financial stability and resilience, particularly for
consumers with uneven or low incomes.
»» Use of transaction AFS remained more common than
use of credit AFS.11
• Consistent with past survey results, AFS use differed
across households. AFS use in 2017 was more common
among lower-income households, less-educated households, younger households, black and Hispanic households, working-age disabled households, and households
with volatile income.
»» Declines in AFS use over time were fairly widespread
across segments of the population.
• AFS use continued to be much higher among unbanked
households than banked households.
»» The proportion of unbanked households that used
AFS, however, decreased substantially from 2013
to 2017. This decrease is attributable to declines in
the use of both transaction and credit AFS over this
period.
Figure ES.9 Alternative Financial Services Use in Past 12
Months by Year, Unbanked Households (Percent)
63.2
57.3
60.5
51.3
54.1
Transaction AFS
2013
2015
12.3
18.6 17.6 16.3
Transaction AFS
2013
10
2015
Figure ES.11 Rates of Saving for Unexpected Expenses or
Emergencies by Banking Status and Year
55.2
2017
6.2
Any AFS
»» The savings rate increased substantially among
Hispanic households from 42.5 percent in 2015 to
48.2 percent in 2017. Moreover, savings rates among
younger households increased more than savings
rates among older households.
Credit AFS
Figure ES.10 Alternative Financial Services Use in Past 12
Months by Year, Banked Households (Percent)
21.7 21.4 20.0
»» As in 2015, rates of saving for unexpected expenses
or emergencies in 2017 were lower among certain
segments of the population, including lower-income
households, less-educated households, older households, black and Hispanic households, and working-age disabled households.
»» Unbanked households continued to save for unexpected expenses or emergencies at a much lower rate than
underbanked and fully banked households.
48.0
16.7 16.5
Any AFS
• Overall, 57.8 percent of households saved for unexpected expenses or emergencies in 2017; that is, they set
aside money in the past 12 months that could be used for
unexpected expenses or emergencies, even if the funds
were later spent. The increase in the savings rate since
2015 (56.3 percent) can be attributed primarily to changes
in income and other characteristics of U.S. households
between 2015 and 2017.
7.0
20.2
56.3
60.0
61.6
17.4
Unbanked
6.5
Underbanked
2015
Fully banked
2017
Credit AFS
2017
Unless noted otherwise, all estimates of AFS use are based on the 12 months before the survey.
For the purposes of this report, transaction AFS include the following nonbank products and services: money orders, check cashing, and international remittances.
Credit AFS include the following nonbank products and services that may be used in lieu of bank credit: payday loans, refund anticipation loans, rent-to-own services,
pawn shop loans, and auto title loans.
11
8 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Figure ES.12 Selected Savings Methods for Households That Saved by Banking Status, 2017 (Percent)
75.0
71.6
66.8
65.1
24.1
23.9
23.7
16.3
10.5
10.1
0.3
All
Savings account
2.0
7.7
2.2
Unbanked
Checking account
0.4
Underbanked
In home, or with family or friends
0.1
Fully banked
Prepaid card
Note: Bars may sum to more than 100 percent because households were asked to select all savings methods used.
• Among all households that saved for unexpected expenses or emergencies in 2017, savings and checking
accounts were the most used savings methods: more
than four in five (85.5 percent) kept savings in one of these
accounts. About one in ten (10.5 percent) households that
saved maintained savings in the home, or with family or
friends.
»» As in 2015, the use of formal (e.g., savings or checking accounts) and informal (e.g., in the home, or with
family or friends) savings methods varied by household
characteristics in 2017.
»» Unbanked households generally saved using informal methods, while underbanked and fully banked
households generally saved using formal methods.
Unbanked households that saved primarily kept
savings in the home, or with family or friends, while
underbanked and fully banked households that saved
primarily used savings accounts.
Credit
Building on the 2015 survey, which introduced questions
about small-dollar bank credit, the 2017 survey included new
questions to capture the full range of credit products that
are likely reported to credit bureaus (i.e., mainstream credit).
Specifically, the 2015 survey asked households whether,
12
in the past 12 months, they had a credit card from Visa,
MasterCard, American Express, or Discover (i.e., credit card)
or a personal loan or line of credit from a bank (i.e., bank
personal loan). Additional questions in the 2017 survey asked
households whether, in the past 12 months, they had a store
credit card; an auto loan; a student loan; a mortgage, home
equity loan, or home equity line of credit (HELOC); or other
personal loans or lines of credit from a company other than
a bank (i.e., other mainstream nonbank).12 A household is
considered to have used mainstream credit if it used any of
the above credit products in the past 12 months.
• Credit cards were the most common type of mainstream
credit (68.7 percent of households had a credit card from
Visa, MasterCard, American Express, or Discover, and
41.6 percent had a store credit card), followed by mortgages, home equity loans, or HELOCs; and auto loans.
»» Use of each mainstream credit product was much
lower among unbanked households, relative to underbanked and fully banked households. For example,
only 7.2 percent of unbanked households had a credit
card, compared with 60.0 percent of underbanked
households and 76.3 percent of fully banked households.
»» Use of mainstream credit products also varied widely
across socioeconomic and demographic groups. In
Other mainstream nonbank credit includes finance company loans and purchase loans or lines of credit from retailers. This category does not include credit AFS.
9
general, lower-income households, less-educated
households, the youngest and oldest households,
black and Hispanic households, and working-age
disabled households were less likely to use most
mainstream credit products.
• Households that did not have mainstream credit in the
past 12 months likely did not have a credit score, which
could make it more difficult to obtain mainstream credit
should a credit need arise.13
»» One in five (19.7 percent) households in 2017 had no
mainstream credit in the past 12 months.
»» Differences in the share of households with no
mainstream credit by banking status were striking.
Four in five (80.2 percent) unbanked households had
no mainstream credit, compared with 21.9 percent
of underbanked households and 14.1 percent of fully
banked households.
»» The share of households with no mainstream credit
also varied substantially across socioeconomic and
demographic groups. Lower-income households,
less-educated households, black and Hispanic households, working-age disabled households, and foreign-born, noncitizen households were more likely not
to have mainstream credit.
»» Differences by race and ethnicity were substantial:
36.0 percent of black households and 31.5 percent
of Hispanic households had no mainstream credit,
compared with 14.4 percent of white households. At
all income levels, black and Hispanic households were
more likely not to have mainstream credit. Racial and
ethnic differences in bank account ownership and socioeconomic and demographic characteristics beyond
income can account for some, but not all, of the racial
and ethnic differences in the likelihood of not having
mainstream credit.
• Two reasons why households may not have mainstream
credit are that they are not interested in having credit or
that they do not appear creditworthy. For the purposes
of this report, we consider a household to have shown
interest in having credit if, in the past 12 months, the
household applied for a credit card or bank personal loan,
thought about applying for a credit card or bank personal
loan but did not because it thought it might be turned
down (i.e., felt discouraged about applying), or use credit
AFS.14
»» Approximately one in six (15.8 percent) households
with no mainstream credit in 2017 showed interest in
having credit.
»» Staying current on bills is one potential indicator of
creditworthiness. About three in four (76.3 percent)
households with no mainstream credit stayed current
on bills in the past 12 months. Among households with
no mainstream credit that showed interest in having
credit, roughly half (46.7 percent) stayed current on
bills. While staying current on bills is an imperfect measure of creditworthiness, it nevertheless provides some
insight into these households’ financial situation.
• Households may use certain credit products, including
credit cards, bank personal loans, and credit AFS, to meet
their small-dollar credit needs. Some households may
have small-dollar credit needs that are not fully met by
mainstream financial institutions. As in the 2015 report,
we classify a household as having unmet demand for
mainstream small-dollar credit if, in the past 12 months,
the household applied for and was denied a credit card
or bank personal loan, felt discouraged about applying, or
used credit AFS.
»» Applying this convention, 12.9 percent of households
had unmet demand for mainstream small-dollar credit
in 2017, compared with 13.7 percent in 2015. The
decline in the share of households with unmet demand
from 2015 to 2017 is consistent with the declines in the
shares of households that used credit AFS or that felt
discouraged about applying for a credit card or bank
personal loan.
»» Among households with unmet demand, 57.2 percent
stayed current on bills in 2017, up slightly from 52.5
percent in 2015.
Households without a credit score may be “credit invisible,” meaning that no one in the household has a record at one of the credit bureaus. Alternatively, a
household member may have a record at one of the credit bureaus but not have sufficient credit history to be scored. At least one active trade line in the past six
months is generally required to generate a credit score.
13
This definition is an approximation and likely does not capture all households that have shown interest in having credit. For example, households may have applied
for or have felt discouraged about applying for other credit products, such as auto loans or student loans.
14
10 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Figure ES.13 Use of Mainstream Credit Products, 2017 (Percent)
Credit card
68.7
Store credit card
41.6
Mortgage, home equity loan, or HELOC
33.8
Auto loan
32.3
Student loan
16.6
Bank personal loan
Other mainstream nonbank
6.9
2.1
No mainstream credit
19.7
Figure ES.14 No Mainstream Credit by Household Race and Ethnicity and Income Level, 2017 (Percent)
Less than $15,000
48.2
$15,000 to $30,000
16.2
$50,000 to $75,000
8.5
3.5
6.8
67.8
48.7
47.9
29.1
$30,000 to $50,000
At least $75,000
63.4
27.9
28.5
18.9
18.5
9.5
Black
Hispanic
White
Note: To simplify the figure, estimates for Asian households and for households of other races and ethnicities are not shown.
11
Table ES.5 Methods Used to Pay Bills and Receive Income in a Typical Month by Banking Status, 2017
For all households that paid bills and received income in a typical month, column percent
All
Unbanked
Underbanked
Fully banked
Electronic payment from bank
68.4
2.5
67.2
73.0
Personal check
61.3
1.2
52.0
67.8
Debit card
47.3
3.1
63.1
45.9
Credit card
24.8
8.4
25.0
25.8
Bank money order
5.7
13.0
11.8
3.5
Cash
15.9
66.1
26.2
9.8
Nonbank money order
6.9
39.1
24.2
0.0
Prepaid card
2.3
22.1
4.0
0.5
Other
1.2
8.0
1.3
0.7
Did not select a method
0.5
3.3
0.3
0.4
Any bank method
93.8
22.7
94.0
98.4
Only bank methods
78.2
6.2
56.6
88.8
Direct deposit into bank account
86.7
5.6
86.6
92.0
Paper check or money order
27.6
45.4
30.8
25.6
Cash
7.9
26.5
10.5
6.0
Direct deposit onto prepaid card
3.4
23.3
5.0
1.6
Other
1.8
10.6
1.9
1.2
Nonbank check casher
1.9
23.7
3.8
0.0
Did not select a method
1.6
10.5
1.3
1.1
Any bank method
93.2
5.6
95.3
98.2
Only bank methods
84.1
2.6
80.0
90.5
A. Paying bills (Percent)
B. Receiving income (Percent)
How Households Conduct Their Financial
Transactions in a Typical Month
To learn more about the extent to which households use bank
and other methods to meet their financial transactions needs,
the 2017 survey asked about the ways households pay bills
and receive income in a typical month.
• From 2015 to 2017, use of paper instruments to handle
these financial transactions declined somewhat, while use
of electronic methods increased.
»» Although personal checks remained the second-most
prevalent method of paying bills, after electronic payments from a bank account, the proportion of households that used personal checks decreased from 2015
to 2017. Over the same period, the proportions that
used electronic payments from a bank account, debit
cards, or credit cards increased.
»» Likewise, the proportion of households that received
income by paper check or money order decreased
from 2015 to 2017, while the proportion that received
income through direct deposit into a bank account
increased.
• As in 2015, unbanked households in 2017 paid bills and
received income primarily using methods outside of the
banking system.
»» Approximately two-thirds paid bills using cash in 2017,
the most prevalent method. Nonbank money orders and
prepaid cards were the next two most prevalent methods
of paying bills.
»» Unbanked households also received income in a variety of ways, but the most prevalent method was paper
check or money order, followed by cash and direct
deposit onto a prepaid card.
• Underbanked households, on the other hand, used banks
extensively to handle their financial transactions. The
key difference between underbanked and fully banked
households is that, in addition to using bank methods,
underbanked households also widely used other methods
to pay bills.
12 | 2017 FDIC National Survey of Unbanked and Underbanked Households
»» Electronic payment from a bank account was the
most prevalent method of paying bills among both
underbanked and fully banked households in 2017.
Relative to the fully banked, use of personal checks
was lower among underbanked households and
use of debit cards was higher. Direct deposit into a
bank account was by far the most prevalent method
of receiving income, both for underbanked and fully
banked households.
»» Approximately one in four underbanked households
used cash to pay bills in a typical month, and a similar
share used nonbank money orders.
Measuring Economic Inclusion
A primary goal of the FDIC National Survey of Unbanked and
Underbanked Households is to assess the inclusiveness of
the U.S. banking system. Specifically, the survey is used to
estimate the proportion of households that do not have an
account at a federally insured depository institution (i.e., the
unbanked rate) and the proportion that have an account but
go outside of the banking system to meet their financial needs
(i.e., the underbanked rate). As consumer financial product
markets evolve and new products mature, measurement of
the unbanked and underbanked may require updating to
reflect such changes and to better assess the inclusiveness
of the banking system.
• In this report and since the survey was first conducted in
2009, a household is categorized as unbanked if no one in
the household has a checking or savings account. General
purpose reloadable prepaid cards that were obtained from
banks may offer many of the same features as checking
accounts as well as a relationship with a retail banking
institution.
»» As a result, unbanked households that use prepaid
cards obtained from banks could be considered
banked. If they were, the unbanked rate in 2017
would fall slightly from 6.5 percent to 6.4 percent.
• In this report and since 2013, a household is classified as
underbanked if it has a checking or savings account and
used one of the following products or services from an
AFS provider in the past 12 months: money orders, check
cashing, international remittances, payday loans, refund
anticipation loans, rent-to-own services, pawn shop loans,
or auto title loans.
»» This underbanked definition does not incorporate
intensity of AFS use: some underbanked households
may routinely use AFS, while others may do so only
sporadically.
»» It also considers a wide range of AFS, including transaction and credit products and services.
• As a result, households categorized as underbanked
in this report are a fairly broad group, with a variety of
experiences and levels of engagement with the banking
system.
»» In 2017, approximately half (48.6 percent) of underbanked households used only bank methods to pay
bills and receive income in a typical month, which we
denote as underbanked group 1.15 The remaining 51.4
percent of underbanked households did not exclusively use bank methods to pay bills and receive income
in a typical month, which we denote as underbanked
group 2.
»» Households in underbanked group 1 were quite
similar to the fully banked in their socioeconomic and
demographic characteristics, savings activity, and
use of mainstream credit products. Compared with
households in underbanked group 1 and with the fully
banked, households in underbanked group 2 had
lower income and educational attainment; were more
likely to be young, black, Hispanic, or working-age
disabled; and were more likely to have volatile income.
»» Use of mobile banking as the primary method of bank
account access was similar across the two underbanked groups. In contrast, use of bank tellers was
more prevalent and use of online banking less prevalent among households in underbanked group 2.
»» Rates of savings for unexpected expenses or emergencies, use of savings or checking accounts for
keeping savings, and use of most mainstream credit products were also lower among households in
underbanked group 2, compared with households in
underbanked group 1 and with the fully banked.
»» Some of the characteristics and behaviors of households in underbanked group 2 were similar to the
characteristics and behaviors of the unbanked, including the share with volatile income, the use of cash to
pay bills or receive income in a typical month, and the
proportion that fell behind on bills.
»» Overall, this analysis suggests that it is important to
consider intensity of transaction AFS use in measuring
the underbanked. If intensity of transaction AFS use
were considered in the classification of underbanked
households, fewer households in underbanked group 1
may be classified as underbanked.
Households in underbanked group 1 were classified as underbanked because either they used credit AFS in the past 12 months, or they used transaction AFS in the
past 12 months but not to pay bills or receive income in a typical month.
15
13
Implications
The survey results presented in this report suggest implications for policymakers, financial institutions, and other stakeholders who are working to improve access to mainstream
financial services.
1. New underwriting technologies could help expand access
to small-dollar credit for banked consumers, including
consumers with little or no credit history. The vast majority
of the 13 percent of households with unmet demand for
mainstream small-dollar credit are banked, and almost all
receive income and pay bills using their bank accounts.
But few of these households applied for a credit card
or bank personal loan. Account balances and transactions may provide information for banks to underwrite
small-dollar credit to some of these households.
2. About one in five households likely have little or no credit
history. The vast majority of these households are banked
and may not seek credit until a need arises. Helping these
households establish and build a credit history may particularly benefit black households, Hispanic households,
and households headed by a working-age individual with
a disability. All of these households are disproportionately
less likely to have mainstream credit.
4. Physical access to bank branches remains important
even as use of mobile banking and online banking has
increased. In 2017, the great majority of banked households visited a bank branch in the past 12 months, and
more than one-third visited ten or more times. In addition,
almost one in six unbanked households visited a bank
branch in the past 12 months. These findings suggest that
branches continue to play an important role for banked
households and that opportunities may exist for branch
staff to inform unbanked households about products and
services that can help meet their financial needs.
5. Unbanked rates for some segments of the population
have declined as economic conditions improved between
2011 and 2017. Still, unbanked rates for these groups,
including black and Hispanic households, remain substantially above the national average. At the same time,
unbanked rates for other population segments, such as
working-age disabled households, have remained high
and stayed fairly constant between 2011 and 2017. Understanding the evolution of unbanked rates for different
population segments and adopting targeted strategies
may help sustain increases in bank account ownership in
future economic downturns and increase access for different population segments with high unbanked rates.
3. Mobile banking holds real promise for deepening the
connection between underbanked households and their
banks while increasing the safety and convenience of
bill payments. A large share of underbanked households
pays bills in a typical month with cash or nonbank money
orders. More than two in five of these households already
use mobile banking to access their bank accounts. Increased use of mobile banking activities by these households may enable them to conduct a greater share of their
basic financial transactions within the banking system.
14 | 2017 FDIC National Survey of Unbanked and Underbanked Households
2017 FDIC National Survey of Unbanked and Underbanked
Households
2. Background and Objectives
Background
When households open an account at a federally insured
depository institution, they establish a mainstream banking
relationship that provides them the opportunity to deposit
funds securely, conduct basic financial transactions, and
accumulate savings.
Despite these benefits, many households—referred to in this
report as “unbanked”—do not have an account at an insured institution. Other households have an account but also
obtained financial products or services from an alternative
financial services (AFS) provider in the past 12 months. These
households are referred to as “underbanked” in this report.
Unbanked and underbanked households present an opportunity for banks to expand access to and utilization of their
products and services.
The FDIC is committed to expanding economic inclusion in
the financial mainstream by ensuring that all Americans have
access to safe, secure, and affordable banking services. The
FDIC National Survey of Unbanked and Underbanked Households is one contribution to this end.
Conducted to assess the inclusiveness of the banking
system, and in partial response to a statutory mandate,
this biennial survey provides estimates of unbanked and
underbanked populations. It also seeks to offer insights that
will inform efforts to better meet the needs of these groups.
The FDIC conducts the household survey in partnership with
the U.S. Census Bureau. Specifically, the FDIC sponsors a
special supplement on unbanked and underbanked households that is administered in conjunction with the Census
Bureau’s Current Population Survey (CPS).
The first FDIC National Survey of Unbanked and Underbanked
Households was conducted in January 2009, and subsequent
surveys were conducted in June 2011, June 2013, June 2015,
and June 2017. Results from these surveys are available at
http://www.economicinclusion.gov.
This report presents the results of the 2017 FDIC National
Survey of Unbanked and Underbanked Households. This
survey was conducted in June 2017 and collected responses
from 35,217 households. See Appendix 1 (FDIC Technical
Notes) for additional details.
Where appropriate, this report discusses trends in survey
results over time. In certain cases, results are not comparable
across years because of changes in the survey instrument.
For example, underbanked rates in 2013, 2015, and 2017
are not comparable to the 2009 or 2011 estimates because
of differences in the types of AFS included in the survey that
were used to categorize households as underbanked.
The results of this survey complement other FDIC efforts
to increase sustainable and safe access to the financial
mainstream. For more information on those efforts and for
additional resources from this survey, including the ability
to query the underlying data, readers should visit
http://www.economicinclusion.gov.
The FDIC encourages researchers, policymakers, consumer
and community groups, and financial institutions to use the
publicly available data to improve understanding of the issues
and challenges unbanked and underbanked households face
when deciding how and where to conduct financial transactions. The information provided in this report, as well as future
analyses produced with the publicly available data, will contribute to efforts to create sustainable banking opportunities
for a broad set of consumers.
What’s New
A number of changes were made to the 2017 survey
instrument to provide additional information about the
characteristics of unbanked and underbanked households.
The details of these changes, summarized below, are
provided in Appendix 2.
The notable additions to the 2017 survey instrument fall into
three main areas.
15
First, to supplement existing questions on use of bank tellers
to access a bank account, the survey included new questions
that asked all households whether they spoke with a teller
or other employee in person at a bank branch in the past 12
months and, if so, how many times. The goal of these questions
is to provide a more complete picture of household use of bank
branches. Banked households may rely on bank branches for
activities other than accessing an account, such as resolving
a problem or asking about products or services. Unbanked
households may visit a bank branch to learn about products or
services or to use those provided to non-account holders.
Second, to supplement existing questions on use of mobile
banking, the survey included a series of questions that asked
households about the ways they used a mobile phone in the
past 12 months for banking activities, such as paying bills,
sending money to others, and depositing a check. Most of
these questions were also asked in the 2013 survey.
Finally, to approximate the share of households that do not
have a credit score, the survey added questions to capture
the full range of credit products that are likely reported to the
major credit bureaus (i.e., mainstream credit). As in 2015,
households were asked whether, in the past 12 months, they
had a credit card from Visa, MasterCard, American Express,
or Discover or whether they had a personal loan or line of
credit from a bank. New questions asked households whether, in the past 12 months, they had a store credit card; an auto
loan; a student loan; a mortgage, home equity loan, or home
equity line of credit; or other personal loans or lines of credit
from a company other than a bank.
16 | 2017 FDIC National Survey of Unbanked and Underbanked Households
2017 FDIC National Survey of Unbanked and Underbanked
Households
3. Banking Status of U.S. Households
2017 National Estimates
Changes in Banking Status
An estimated 6.5 percent of U.S. households were “unbanked” in 2017, meaning that no one in the household had a
checking or savings account (see Figure 3.1). This proportion
represents approximately 8.4 million U.S. households composed of 14.1 million adults and 6.4 million children.16
The proportion of U.S. households that were unbanked (i.e.,
the unbanked rate) in 2017—6.5 percent—declined to the
lowest level since the survey began in 2009, as shown in
Figure 3.2. Since the survey was last administered in 2015,
the unbanked rate has fallen by 0.5 percentage points.17
An additional 18.7 percent of U.S. households (24.2 million)
were “underbanked” in 2017, meaning that the household had
a checking or savings account and used one of the following
products or services from an alternative financial services
(AFS) provider in the past 12 months: money orders, check
cashing, international remittances, payday loans, refund anticipation loans, rent-to-own services, pawn shop loans, or auto
title loans. These underbanked households were made up of
48.9 million adults and 15.4 million children.
The decline in the unbanked rate from 2015 to 2017 can be
explained almost entirely by changes in household characteristics across survey years, particularly improvements in
the socioeconomic circumstances of U.S. households. After
accounting for these changes, the remaining difference in the
unbanked rate from 2015 to 2017 was very close to zero and
no longer statistically significant.18
Most U.S. households (68.4 percent) were “fully banked” in
2017, meaning that the household had a bank account and
did not use AFS in the past 12 months. The remaining 6.3
percent of U.S. households had a bank account, but information on their use of AFS was insufficient to categorize
the household as either underbanked or fully banked (i.e.,
banked, underbanked status unknown).
Figure 3.1 Banking Status of U.S. Households, 2017
(Percent)
Unbanked
6.5
Underbanked
18.7
Fully
banked
68.4
Banked,
underbanked
status unknown
6.3
Figure 3.2 National Estimates, Household Unbanked
Rates by Year
7.6
2009
8.2
2011
7.7
2013
7.0
2015
6.5
2017
The proportion of households that were underbanked (i.e.,
the underbanked rate) fell from 19.9 percent in 2015 to 18.7
percent in 2017, as shown in Table 3.1. This decline was
attributable in part to changes in household characteristics
between 2015 and 2017, particularly improvements in the
socioeconomic circumstances of U.S. households. Even
after accounting for these changes, the remaining decline
in the underbanked rate from 2015 to 2017 was statistically
significant.
Adults are defined as people aged 16 and older. The estimates of 14.1 million adults and 6.4 million children may understate the total number of people in the United
States who do not have access to a bank account because these figures do not include residents of “banked” households who do not have an account in their name
and do not benefit from a bank account owned by another household resident.
16
All differences discussed in the text of this report are statistically significant at the 10 percent level, unless noted otherwise. In other words, there is a 10 percent or
lower probability that the difference observed in the survey is due to chance.
17
A linear probability model was estimated to account for changes from 2015 to 2017 in the distribution of households across the household characteristics listed in
Appendix Table A.2. Changes in the socioeconomic characteristics of households (annual income level, monthly income volatility, employment status, homeownership
status, and educational attainment) between 2015 and 2017 accounted for almost all of the difference in unbanked rates between 2015 and 2017. Adding controls for
the remaining demographic characteristics listed in Appendix Table A.2 had little effect on the remaining difference.
18
17
Comparing the fully banked rate in 2017 to earlier years is
made more difficult by an increase in the proportion of households that were banked but provided insufficient information
on their use of AFS to be categorized as either underbanked
or fully banked. Table 3.1 shows that the proportion of U.S.
households that were fully banked in 2017 (68.4 percent) was
slightly higher than the 2015 estimate (68.0 percent).19
Banking Status by Household Characteristics
Consistent with previous surveys, banking status in 2017
varied considerably across the U.S. population. For example,
unbanked and underbanked rates were higher among lower-income households, less-educated households, younger
households, black and Hispanic households, working-age
disabled households, and households with volatile income.20
Unbanked rates in 2017 were lower than or similar to
unbanked rates in recent years for most segments of the
population, as illustrated in Table 3.2. For example, recent
declines in unbanked rates have been particularly sharp for
younger households. Among households aged 15 to 24, the
unbanked rate in 2017 was 10.0 percent, down from 13.1
percent in 2015 and 15.7 percent in 2013. Declines were also
substantial for households aged 25 to 34.
Unbanked rates among black and Hispanic households have
also sharply declined in recent years. Specifically, 16.9 percent of black households were unbanked in 2017, down from
18.2 percent in 2015 and 20.6 percent in 2013. 21 Among
Hispanic households, 14.0 percent were unbanked in 2017,
down from 16.2 percent in 2015 and 17.9 percent in 2013.
Further, unbanked rates declined for Asian households from
2015 to 2017, reversing most of the increase in the unbanked
rate among this group from 2013 to 2015.
Despite these improvements, unbanked rates for younger
households and for black and Hispanic households remained
substantially higher than the overall unbanked rate in 2017.
Unbanked rates did not decline in recent years for a few segments of the population. For example, among working-age
disabled households, unbanked rates were similar in 2013,
2015, and 2017.
Reflecting the decline in the underbanked rate at the national level between 2015 and 2017, underbanked rates also
declined for many segments of the population during that period. For example, as shown in Table 3.3, underbanked rates
decreased for households with less than $15,000 in income,
households with a high school diploma (but no college), and
working-age disabled households.
Although underbanked rates among certain groups declined
considerably from 2015 to 2017, the change relative to 2013
was not quite as large in many cases. For example, among
households with less than $15,000 in income, the underbanked rate in 2017 (20.9 percent) was 3.4 percentage points
lower than in 2015 (24.3 percent) but only 1.4 percentage
points lower than in 2013 (22.4 percent). (See Appendix
Table A.4.)
Table 3.1 National Estimates, Household Banking Status by Year
For all households, row percent
Year
Number of Households
(1000s)
Unbanked
(Percent)
Underbanked
(Percent)
Fully banked
(Percent)
Banked, underbanked
status unknown
(Percent)
2013
123,750
7.7
20.0
67.0
5.3
2015
127,538
7.0
19.9
68.0
5.0
2017
129,276
6.5
18.7
68.4
6.3
Excluding banked households with unknown underbanked status, the fully banked rate increased from 71.6 percent in 2015 to 73.1 percent in 2017, and this increase
was statistically significant. The decline in the underbanked rate from 2015 to 2017 was statistically significant regardless of whether banked households with unknown
underbanked status were excluded (excluding such households, the underbanked rate was 21.0 percent in 2015 and 20.0 percent in 2017).
19
For characteristics that vary at the person-level, such as race, age, and education, the characteristics of the owner or renter of the home (i.e., the householder) are
used to represent the household. For convenience, abbreviated language is used when referring to certain household characteristics. For example, the term “white
household” refers to a household for which the householder has been identified as white, non-black, non-Hispanic, and non-Asian. The phrase “working-age disabled
household” refers to a household in which the householder has a disability and is aged 25 to 64. See Appendix 1 for additional details. For monthly income volatility,
the 2015 and 2017 surveys asked households whether their income over the past 12 months “was about the same each month,” “varied somewhat from month to
month,” or “varied a lot from month to month.” The term “volatile income” refers to a household with income that varied somewhat or a lot from month to month.
20
The decline in the unbanked rate for black households from 2013 to 2017 was no longer statistically significant after accounting for changes in the other household
characteristics listed in Appendix Table A.2 (except for monthly income volatility, which is not available for 2013). Most of the decline can be attributed to changes in
income and the other household characteristics across survey years.
21
18 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Table 3.2 Unbanked Rates by Selected Household Characteristics and Year
For all households
2013
(Percent)
2015
(Percent)
2017
(Percent)
Difference
(2017–2015)
7.7
7.0
6.5
–0.5*
Less than $15,000
27.7
25.6
25.7
0.1
$15,000 to $30,000
11.4
11.8
12.3
0.4
$30,000 to $50,000
5.1
5.0
5.1
0.1
$50,000 to $75,000
1.7
1.6
1.5
–0.1
At least $75,000
0.5
0.5
0.6
0.1
No high school diploma
25.1
23.2
22.4
–0.8
High school diploma
10.8
9.7
9.4
–0.2
Some college
5.6
5.5
5.1
–0.4
College degree
1.1
1.1
1.3
0.1
15 to 24 years
15.7
13.1
10.0
–3.1*
25 to 34 years
12.5
10.6
8.5
–2.1*
35 to 44 years
9.0
8.9
7.8
–1.2*
45 to 54 years
7.5
6.7
6.9
0.2
55 to 64 years
5.6
5.8
5.9
0.1
65 years or more
3.5
3.1
3.9
0.8*
Black
20.6
18.2
16.9
–1.3
Hispanic
17.9
16.2
14.0
–2.3*
Asian
2.2
4.0
2.5
–1.5*
White
3.6
3.1
3.0
–0.1
Other
15.0
11.1
12.8
1.7
Disabled, age 25 to 64
18.4
17.6
18.1
0.5
Not disabled, age 25 to 64
7.2
6.5
5.7
–0.8*
Income was about the
same each month
5.7
5.6
–0.1
Income varied somewhat
from month to month
8.7
6.8
–1.9*
Income varied a lot from
month to month
12.9
13.2
0.3
Characteristics
All
Family income
Education
Age group
Race/Ethnicity
Disability status
Monthly income
volatility
Notes: Monthly income volatility is not available for 2013. Asterisks indicate differences that are statistically significant at the 10 percent level. See Appendix Table
A.3 for estimates by other household characteristics and for selected confidence intervals.
19
Table 3.3 Underbanked and Fully Banked Rates by Selected Household Characteristics and Year
For all households
Underbanked
Characteristics
Banked,
underbanked status unknown
Fully banked
2015
(Percent)
2017
(Percent)
Difference
(2017–2015)
2015
(Percent)
2017
(Percent)
Difference
(2017–2015)
2015
(Percent)
2017
(Percent)
Difference
(2017–2015)
19.9
18.7
–1.2*
68.0
68.4
0.4
5.0
6.3
1.3*
Less than $15,000
24.3
20.9
–3.4*
45.1
47.7
2.6*
4.9
5.7
0.7
$15,000 to $30,000
23.6
22.4
–1.2
59.5
58.3
–1.1
5.1
7.0
1.9*
$30,000 to $50,000
23.7
22.8
–0.9
66.2
65.4
–0.8
5.1
6.8
1.7*
$50,000 to $75,000
20.2
19.7
–0.5
73.0
72.8
–0.2
5.1
6.0
0.9*
At least $75,000
13.4
13.3
–0.1
81.3
79.9
–1.3*
4.9
6.2
1.3*
No high school
diploma
25.9
24.3
–1.6
46.4
46.3
–0.1
4.5
7.0
2.5*
High school diploma
22.2
20.3
–1.8*
62.9
63.7
0.8
5.3
6.5
1.2*
Some college
22.0
20.8
–1.2*
67.7
67.8
0.1
4.8
6.3
1.5*
College degree
14.5
14.4
–0.1
79.1
78.3
–0.8
5.2
6.1
0.9*
15 to 24 years
29.4
29.3
–0.1
52.1
56.5
4.5*
5.5
4.2
–1.3
25 to 34 years
24.5
23.1
–1.4
60.8
62.5
1.6
4.0
5.9
1.9*
35 to 44 years
22.7
22.2
–0.5
63.1
63.6
0.5
5.3
6.5
1.2*
45 to 54 years
21.1
19.3
–1.8*
67.5
67.1
–0.3
4.8
6.7
1.9*
55 to 64 years
18.5
17.8
–0.8
70.9
70.3
–0.6
4.8
6.0
1.2*
65 years or more
13.0
11.6
–1.4*
78.1
77.5
–0.6
5.8
7.0
1.2*
Black
31.1
30.4
–0.7
45.5
45.8
0.3
5.2
6.9
1.7*
Hispanic
29.3
28.9
–0.3
48.9
49.7
0.8
5.6
7.4
1.8*
Asian
21.0
17.5
–3.5*
67.2
69.2
2.0
7.8
10.8
2.9*
White
15.6
14.1
–1.5*
76.6
77.1
0.5
4.7
5.7
1.1*
Other
27.5
28.0
0.5
56.7
55.8
–0.9
4.6
3.3
–1.3
Disabled, age 25 to 64
28.4
24.7
–3.7*
49.7
52.2
2.5
4.3
5.0
0.8
Not disabled, age 25
to 64
20.6
19.9
–0.7
68.1
68.0
–0.1
4.8
6.4
1.6*
Income was about the
same each month
19.1
18.1
–1.0*
74.4
75.2
0.8*
0.8
1.0
0.2*
Income varied
somewhat from
month to month
26.6
26.2
–0.5
64.0
66.1
2.1*
0.6
0.9
0.3
Income varied a lot
from month to month
30.9
28.2
–2.7
55.2
57.9
2.7
1.0
0.7
–0.3
All
Family income
Education
Age group
Race/Ethnicity
Disability status
Monthly income
volatility
Notes: Asterisks indicate differences that are statistically significant at the 10 percent level. See Appendix Tables A.4 and A.5 for underbanked and fully banked rates
by other household characteristics and for selected confidence intervals.
20 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Figure 3.3 Unbanked Rates by State, 2017
WA
ND
MT
OR
MN
ID
MI
WY
IL
UT
CO
KS
CA
WV
KY
TN
OK
NM
AZ
MO
PA
OH
IN
TX
MD
VA
MA
CT
NJ
RI
DE
DC
NC
AR
MS
VT
NH
NY
IA
NE
NV
ME
WI
SD
SC
AL
GA
LA
Less than 3.4
3.4 to 5.0
5.0 to 6.5
FL
AK
6.5 to 7.8
HI
At least 7.8
Figure 3.4 Underbanked Rates by State, 2017
WA
ND
MT
OR
MN
SD
ID
MI
WY
NV
IL
CO
KS
CA
AZ
OK
NM
MO
WV
KY
TN
LA
AL
MD
VA
MA
CT
NJ
RI
DE
DC
NC
AR
MS
TX
PA
OH
IN
VT
NH
NY
IA
NE
UT
ME
WI
SC
GA
Less than 15.7
15.7 to 17.9
FL
AK
HI
17.9 to 19.3
19.3 to 21.4
At least 21.4
21
Banking Status by Geography
Regional variation in unbanked and underbanked rates in
2017 was similar to previous years: unbanked and underbanked rates were highest in the South. The unbanked rate
in the South was 7.7 percent in 2017, compared with 5.4
percent in the Midwest and 6.0 percent in the Northeast and
West. However, the gaps in unbanked rates between the
South and the other regions have narrowed since 2015. The
unbanked rate for the South in 2017 was 1.0 percentage point
lower than the 2015 estimate (8.7 percent), while the other
regions experienced smaller changes in unbanked rates for
that period. The South also saw a slight decline in the underbanked rate from 2015 to 2017, but the declines were more
pronounced for the other regions. (See Appendix Tables A.3
and A.4.)
Unbanked and underbanked rates in 2017 varied widely
across states, as illustrated in Figures 3.3 and 3.4. Reflecting
the regional variation described above and similar to estimates from previous years, unbanked and underbanked rates
were generally higher among states in the South. Unbanked
rates ranged from 1.5 percent (Vermont and Minnesota) to
15.8 percent (Mississippi), while underbanked rates ranged
from 11.6 percent (Vermont and Wisconsin) to 25.1 percent
(Nevada). Some states saw large changes in unbanked rates
in recent years. For example, the unbanked rate in Arizona
was 5.4 percent in 2017, down from 8.5 percent in 2015 and
12.8 percent in 2013, while the unbanked rate in South Dakota was 8.1 percent in 2017, up from 4.2 percent in 2015 and
2013. (See Appendix Tables A.7 – A.14 for detailed estimates
by state and metropolitan statistical area [MSA] and for selected confidence intervals.22)
Transitions in Bank Account Ownership
As discussed in previous reports, bank account ownership is
not static and some households appear to cycle in and out
of the banking system. Table 3.4 segments households by
changes in bank account ownership within the past year. 23
In 2017, 5.8 percent of households were longer-term unbanked, meaning that they did not have a bank account at
the time of the survey or at any time in the 12 months before
the survey.24 A small proportion of households, 0.6 percent,
were recently unbanked, meaning that they did not have an
account at the time of the survey but did at some point in the
12 months before.
Another 3.9 percent of households were recently banked,
meaning that they had an account at the time of the survey
but did not at some point in the 12 months before the survey.
The remaining 89.7 percent of households were longer-term
banked, meaning that they had an account at the time of the
survey and continually during the 12 months before.
These patterns are fairly similar to 2015 and 2013. One notable difference is that while similar proportions of households
were recently banked in 2015 and 2017, the proportion more
than doubled from 2013 to 2015.25
Table 3.4 Household Banking Status Transitions by Year
For all households, row percent
22
Year
Number of
Households
(1000s)
Longer–term
unbanked
(Percent)
Recently
unbanked
(Percent)
Recently
banked
(Percent)
Longer–term
banked
(Percent)
2013
120,918
7.1
0.7
1.6
90.7
2015
125,402
6.2
0.8
3.8
89.2
2017
127,085
5.8
0.6
3.9
89.7
See http://www.economicinclusion.gov/five-year for five-year estimates of unbanked and underbanked rates at the state and MSA levels and for confidence intervals.
The analysis of household banking status transitions excludes 543 observations (representing roughly 2.2 million households) with missing information on recent
banking status. The 2013 estimates in Table 3.4 differ from those published in the 2013 report because the 2013 report also dropped observations with missing
information on life events that may have contributed to household transitions into and out of the banking system, questions about which were not repeated in later
surveys.
23
Households that were longer-term unbanked may never have had an account, or they may have had an account at some point more than 12 months before the survey.
24
The increase from 2013 to 2015 in the proportion of households that were recently banked remained large and statistically significant even after accounting for
changes in the household characteristics listed in Appendix Table A.2 (except for monthly income volatility, which is not available for 2013).
25
22 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Unbanked Households: Previous Banking Status
and Future Banking Plans
Looking beyond one-year transitions in bank account ownership further supports the notion that bank account ownership
is dynamic. As shown in Table 3.5, nearly half (47.0 percent)
of unbanked households in 2017 had a bank account at some
point in the past (i.e., were previously banked), similar to
previous years.26
Table 3.5 Previous Banking Status of Unbanked
Households by Year
For all unbanked households, row percent
Year
Number of
Unbanked
Households
(1000s)
Once had
bank account
(Percent)
Never had
bank account
(Percent)
2013
9,437
46.6
53.4
2015
8,811
47.3
52.7
2017
8,207
47.0
53.0
The proportion of unbanked households that were “very
likely” or “somewhat likely” to open an account in the next
12 months declined in 2017 compared with earlier years,
while the proportion that were “not at all likely” increased. As
shown in Table 3.6, 25.0 percent of unbanked households in
2017 were very likely or somewhat likely to open an account,
down from 37.9 percent in 2013.27 Further, 58.7 percent of
unbanked households in 2017 were not at all likely to open
an account, up from 40.0 percent in 2013.28 This increase
was fairly widespread among segments of the unbanked
population. (See Appendix Table A.15.)
Reasons Households Were Unbanked
As in previous years, the 2017 survey asked unbanked
households about the reasons why they did not have a bank
account. Findings are similar to those reported in previous years.
keep in an account” as a reason for not having an account,
the most commonly cited reason. This reason was also the
most commonly cited main reason for not having an account
(34.0 percent). Almost one-third (30.2 percent) of unbanked
households cited “Don’t trust banks” as a reason for not
having an account, the second-most commonly cited reason.
This reason was also the second-most commonly cited main
reason (12.6 percent).
Other commonly cited reasons were “Avoiding a bank gives
more privacy,” “Bank account fees are too high,” and “Bank
account fees are unpredictable.” Among these reasons,
“Bank account fees are too high” was cited as a main reason
more often than “Avoiding a bank gives more privacy” and
“Bank account fees are unpredictable.”
Reasons for not having an account were generally similar
across unbanked households in 2017, regardless of whether
they previously had an account or were likely to open an
account in the future. A few exceptions are worth noting. As
in previous years, higher proportions of previously banked
households cited “Bank account fees are too high” (29.9
percent) or “Bank account fees are unpredictable” (24.9
percent) in 2017, compared with households that never had
an account (21.1 and 17.0 percent, respectively). Moreover,
higher proportions of unbanked households that were not
at all likely or not very likely to open a bank account in the
next 12 months cited “Don’t trust banks” (36.2 and 31.5
percent, respectively) in 2017, compared with unbanked
households that were somewhat likely or very likely to open a
bank account in the next 12 months (24.7 and 21.0 percent,
respectively). (See Appendix Tables A.16 – A.19 for cited and
main reasons for not having an account by previous banking
status and the likelihood of opening an account.)
As illustrated in Figure 3.5, more than half (52.7 percent) of
unbanked households cited “Do not have enough money to
The analysis of previous banking status excludes 50 observations (representing roughly 0.2 million unbanked households) with missing information on previous
banking status. The 2013 and 2015 estimates in Table 3.5 differ from those published in earlier reports because such observations were not dropped in earlier reports.
26
The analysis of future banking plans excludes 164 observations (representing roughly 0.8 million unbanked households) with missing information on the likelihood
of opening a bank account in the next 12 months. The 2013 and 2015 estimates in Table 3.6 differ from those published in earlier reports because such observations
were not dropped in earlier reports.
27
The proportion of unbanked households that were not at all likely to open an account in the next 12 months was substantially higher in 2017 than in 2013, even after
accounting for changes in the household characteristics listed in Appendix Table A.2 (except for monthly income volatility, which is not available for 2013) and in the
use of prepaid cards between 2013 and 2017.
28
23
Table 3.6 Unbanked Households’ Likelihood of Opening a Bank Account in Next 12 Months by Year
For all unbanked households, row percent
Year
Number of
Unbanked
Households
(1000s)
Very likely
(Percent)
Somewhat likely
(Percent)
Not very likely
(Percent)
Not at all likely
(Percent)
2013
9,021
14.6
23.3
22.1
40.0
2015
8,358
10.2
18.2
19.4
52.2
2017
7,682
9.5
15.6
16.3
58.7
Figure 3.5 Reasons for Not Having a Bank Account, Unbanked Households, 2017 (Percent)
Do not have enough money to keep in account
34.0
Don't trust banks
Avoiding bank gives more privacy
Inconvenient hours
Inconvenient locations
Other reason
24.7
8.6
20.2
1.3
ID, credit, or former bank account problems
Banks do not offer needed products or services
28.2
3.0
Account fees too high
Account fees unpredictable
30.2
12.6
14.0
4.9
13.1
1.6
4.1
2.1
9.4
9.2
14.9
12.3
Cited
24 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Main
52.7
2017 FDIC National Survey of Unbanked and Underbanked
Households
4. Banked Households: Types of Accounts, Methods Used to Access Accounts, and Bank Branch Visits
Types of Accounts Owned by Banked Households
Savings and checking account ownership among banked
households in 2017 was similar to previous years, as shown
in Table 4.1.29 Almost all banked households had a checking
account (98.2 percent), while roughly three in four (78.0
percent) had a savings account.
Savings account ownership rates in 2017 varied widely
across the population. Differences by income and education
were especially pronounced. For example, among banked
households with less than $15,000 in income, only 52.8
percent had a savings account in 2017. In contrast, 91.0
percent of households with income of $75,000 or more had
a savings account in 2017. In addition, savings account
ownership rates were lower among Hispanic households,
working-age disabled households, and households in rural
areas. (See Appendix Table B.2 for details.)
Methods Banked Households Used to Access Their
Accounts
Knowing how households access their bank accounts can
help inform discussions about how best to serve different
groups of consumers. As in the 2013 and 2015 surveys,
banked households were asked about the methods they used
to access their accounts in the past 12 months and about the
primary method used.30 The results show that use of mobile
banking continued to increase sharply, while use of bank
tellers declined. Use of bank tellers, however, remained quite
prevalent, particularly among segments of the population that
had higher unbanked and underbanked rates.
As shown in Table 4.2, in 2017, almost three in four (73.6
percent) banked households used bank tellers to access their
accounts in the past 12 months, a higher proportion than any
other method asked about in the survey.31 However, use of
bank tellers declined modestly between 2013 and 2017, while
use of mobile and online banking increased. The growth in
the use of mobile banking was particularly striking, rising from
23.2 percent in 2013 to 31.9 percent in 2015 and 40.4 percent
in 2017.
Table 4.3 shows the primary method banked households used
to access their accounts. In 2017, online banking remained
the most prevalent primary method of account access (36.0
percent), despite having declined slightly from 2015. Use of
mobile banking increased sharply from 2013 to 2017, while
use of bank tellers declined substantially. Even with the
decline in the use of bank tellers, this method remained the
second-most prevalent primary method.
Table 4.1 Types of Accounts Owned by Banked Households by Year
For all banked households, row percent
Year
Number of
Households
(1000s)
Checking and
savings
(Percent)
Savings only
(Percent)
Checking only
(Percent)
Memo:
Has savings
(Percent)
Memo:
Has checking
(Percent)
2013
111,926
73.8
2.2
24.0
76.0
97.8
2015
116,137
75.8
2.0
22.2
77.8
98.0
2017
118,253
76.2
1.8
22.0
78.0
98.2
As in previous years, the 2017 survey asked about savings and checking account ownership for each person within the household. The analysis of checking and
savings account ownership presented in this section excludes 632 observations (representing roughly 2.6 million banked households) where at least one person in
the household had missing information on bank account type, and there was not enough information from the remaining persons in the household to categorize the
household by the types of bank accounts owned. Estimates of checking and savings account ownership among banked households presented in this section may
differ slightly from the 2013 report because observations with missing information on bank account type were not dropped in the 2013 report.
29
Specifically, banked households were asked about bank tellers, ATMs/kiosks, telephone banking, online banking, mobile banking, and other methods of account
access used in the past 12 months. Households were then asked which method was their primary (i.e., most common) method used. The primary method of account
access does not necessarily reflect intensity of use.
30
The analysis of bank account access methods presented in this section excludes 1,503 observations (representing roughly 5.8 million banked households) that did
not access their accounts in the past 12 months or that did not report whether they accessed their accounts.
31
25
Table 4.4 shows changes between 2015 and 2017 in the
shares of banked households that used bank tellers, online
banking, or mobile banking as their primary method of
account access by banking status and selected household
characteristics. Changes from 2015 to 2017 within segments
of the population were generally similar to overall patterns:
use of bank tellers declined as did use of online banking—
though not by as much as use of bank tellers—and use of
mobile banking increased considerably.
Changes in primary account access methods from 2015 to
2017 differed across age groups. Among older and younger
households, use of bank tellers declined and use of mobile
banking increased, consistent with overall trends. However,
use of online banking increased among older households but
declined among younger households. For the youngest age
group, mobile banking is now the most prevalent primary
method of account access. More than one-third (36.1 percent)
of banked households aged 15 to 24 used mobile banking as
their primary method, compared with 26.2 percent that used
online banking as their primary method.
Estimates show that use of mobile banking grew substantially
among both underbanked and fully banked households. As
in 2015, more underbanked households than fully banked
households used mobile banking as their primary account
access method in 2017 (20.8 percent compared with 14.5
percent). Declines in the use of bank tellers were more
pronounced for underbanked households than for fully
banked households. In 2017, the proportion of underbanked
households that primarily used bank tellers fell by 5.5
percentage points to 22.3 percent, while the proportion
among fully banked households fell by 3.5 percentage
points to 24.6 percent. For underbanked and fully banked
households, use of online banking as the primary method of
account access changed very little from 2015 to 2017. The
proportion of underbanked households that primarily used
online banking (26.8 percent in 2017) remained lower than the
proportion among fully banked households (38.9 percent).
As in prior surveys, use of bank tellers as the primary
means of account access remained quite prevalent among
certain segments of the population, including lower-income
households, less-educated households, older households,
and households in rural areas. These groups were also
disproportionately more likely to access their accounts using
only bank tellers. For example, nearly one-third of banked
households with no high school diploma and about one in
five banked households in rural areas exclusively used bank
tellers to access their accounts in 2017. Overall, 12.6 percent
of banked households accessed their accounts using only
bank tellers in 2017, compared with 14.6 percent in 2015.
(See Appendix Table B.10 for details.)
Table 4.2 All Methods Used to Access Bank Accounts by Year
For all banked households that accessed their account in the past 12 months, row percent
Year
Number of
Households
(1000s)
Bank teller
(Percent)
ATM/Kiosk
(Percent)
Telephone
banking
(Percent)
Online
banking
(Percent)
Mobile
banking
(Percent)
Other
(Percent)
2013
108,295
78.8
69.6
26.1
55.1
23.2
1.0
2015
113,315
75.5
69.8
27.0
60.4
31.9
1.1
2017
115,040
73.6
71.6
28.9
63.0
40.4
0.9
Note: Row percentages sum to more than 100 because households were asked to select all bank account access methods used.
Table 4.3 Primary Method Used to Access Bank Accounts by Year
For all banked households that accessed their account in the past 12 months, row percent
Year
Number of
Households
(1000s)
Bank teller
(Percent)
ATM/Kiosk
(Percent)
Telephone
banking
(Percent)
Online
banking
(Percent)
Mobile
banking
(Percent)
Other
(Percent)
2013
108,295
32.2
24.4
3.3
32.9
5.7
0.8
2015
113,315
28.2
21.0
3.0
36.9
9.5
0.9
2017
115,040
24.3
19.9
2.9
36.0
15.6
0.7
26 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Table 4.4 Use of Bank Tellers, Online Banking, or Mobile Banking as Primary Method of Account Access by Banking
Status and Selected Household Characteristics and Year
For all banked households that accessed their account in the past 12 months
Bank teller
Characteristics
All
Online banking
2015
(Percent)
2017
(Percent)
Difference
(2017-2015)
28.2
24.3
-3.8*
2015
2017
(Percent) (Percent)
36.9
36.0
Mobile banking
Difference
(2017-2015)
2015
2017
Difference
(Percent) (Percent) (2017-2015)
-0.9*
9.5
15.6
6.1*
Banking status
Underbanked
27.8
22.3
-5.5*
27.6
26.8
-0.8
12.6
20.8
8.2*
Fully banked
28.2
24.6
-3.5*
39.9
38.9
-1.0*
8.7
14.5
5.8*
Family income
Less than $15,000
41.7
38.8
-2.8*
18.0
17.2
-0.8
7.1
11.2
4.2*
$15,000 to $30,000
40.5
38.0
-2.5*
20.8
19.4
-1.3
8.1
11.7
3.6*
$30,000 to $50,000
32.5
28.9
-3.6*
29.1
27.7
-1.4
9.7
16.0
6.3*
$50,000 to $75,000
25.8
23.3
-2.5*
39.7
38.0
-1.6
11.3
15.8
4.5*
At least $75,000
16.7
13.3
-3.4*
53.6
50.6
-3.0*
9.7
17.9
8.2*
50.8
46.2
-4.5*
11.8
10.8
-1.0
4.0
8.2
4.2*
Education
No high school diploma
High school diploma
38.2
33.8
-4.4*
24.5
24.7
0.2
7.5
11.6
4.1*
Some college
25.6
22.9
-2.6*
36.8
35.0
-1.7*
11.6
17.5
5.9*
College degree
17.9
14.8
-3.1*
51.5
49.1
-2.4*
10.4
18.2
7.8*
15.9
13.3
-2.7*
31.4
26.2
-5.2*
25.0
36.1
11.1*
13.1*
Age group
15 to 24 years
25 to 34 years
14.3
10.6
-3.6*
42.6
35.7
-6.9*
21.9
35.0
35 to 44 years
16.9
13.6
-3.3*
45.8
42.4
-3.4*
14.3
22.6
8.4*
45 to 54 years
22.9
18.7
-4.2*
42.0
42.6
0.6
7.6
13.2
5.7*
55 to 64 years
31.7
26.1
-5.6*
37.3
39.0
1.6*
3.4
7.0
3.7*
65 years or more
48.7
45.1
-3.6*
23.8
26.9
3.1*
1.2
2.7
1.4*
Race/Ethnicity
Black
30.1
23.8
-6.3*
25.1
24.3
-0.8
11.3
17.7
6.4*
Hispanic
29.3
25.9
-3.4*
27.2
25.8
-1.5
12.6
19.3
6.7*
Asian
25.5
19.1
-6.4*
44.4
46.0
1.7
9.0
16.3
7.3*
White
27.9
24.5
-3.3*
40.0
39.2
-0.8
8.6
14.5
5.8*
Other
25.4
24.5
-0.8
33.8
29.8
-4.0
12.5
18.5
6.0*
Disability status
Disabled, age 25 to 64
32.4
28.7
-3.7*
25.9
26.5
0.6
6.6
10.0
3.5*
Not disabled, age 25 to 64
20.6
16.3
-4.3*
43.8
41.6
-2.2*
11.9
19.8
8.0*
Income was about the same
each month
28.6
24.8
-3.8*
37.3
36.6
-0.7
8.7
14.9
6.2*
Income varied somewhat
from month to month
25.3
20.8
-4.5*
37.4
35.2
-2.1*
13.0
19.6
6.6*
Income varied a lot from
month to month
30.2
24.9
-5.3*
33.7
35.8
2.1
13.0
19.3
6.3*
24.6
19.8
-4.8*
36.9
35.9
-1.0
11.1
18.1
7.1*
Monthly income volatility
Metropolitan status
Metropolitan area - principal
city
Metropolitan area - balance
24.9
21.8
-3.2*
40.8
39.4
-1.3*
9.5
15.6
6.1*
Not in metropolitan area
41.5
37.8
-3.8*
27.4
27.4
-0.1
6.7
11.2
4.5*
Not identified
31.9
28.8
-3.1*
34.0
33.5
-0.5
8.9
14.5
5.6*
Notes: Asterisks indicate differences that are statistically significant at the 10 percent level. See Appendix Tables B.5 – B.9 for estimates by other household
characteristics and for selected confidence intervals.
27
Mobile Phone, Smartphone, and Home Internet Access
As noted in earlier survey reports, financial institutions—banks and nonbanks—are seeking to interact with their customers
through the internet and mobile phones, especially smartphones.
Table 4.5 shows that smartphone access grew markedly from 2013 to 2017. As in 2013 and 2015, more than eight in ten
households had access to a mobile phone in 2017.
Mobile phone and smartphone access continued to be lower among unbanked households than underbanked and fully
banked households. However, approximately half (49.5 percent) of unbanked households owned or had regular access to a
smartphone in 2017, a substantial increase from 2013 and 2015.
As in 2015, unbanked households in 2017 were much more likely to own or have regular access to a smartphone (49.5 percent) than to have internet access at home using a desktop, laptop, or tablet computer (28.5 percent). Among mobile phone,
smartphone, and home internet access, the largest difference between unbanked households and underbanked and fully
banked households continued to be in home internet access.
Table 4.5 Mobile Phone, Smartphone, and Home Internet Access by Banking Status and Year
For all households
Mobile phone
Smartphone
Internet at home
2013
2015
2017
Difference
2013
2015
2017
Difference
2015
2017
Difference
(Percent) (Percent) (Percent) (2017-2015) (Percent) (Percent) (Percent) (2017-2015) (Percent) (Percent) (2017-2015)
All
82.7
84.2
84.5
0.3
55.7
67.1
72.7
5.6*
72.0
72.6
Unbanked
68.1
69.0
Underbanked
90.5
91.4
Fully banked
86.8
88.6
0.6
70.5
1.5
33.1
42.9
49.5
6.6*
27.7
28.5
0.8
92.8
1.4*
64.5
75.5
83.2
7.7*
72.8
76.1
3.2*
90.0
1.4*
59.0
71.1
77.6
6.5*
80.6
81.4
0.8*
Banking status
Notes: Asterisks indicate differences that are statistically significant at the 10 percent level. Estimates of internet access at home are not available for 2013.
See Appendix Tables B.15 – B.17 for estimates by household characteristics and for selected confidence intervals.
Bank Branch Visits Among Banked Households
The 2017 survey included new questions that asked all
households whether they spoke with a teller or other employee in person at a bank branch (i.e., visited a bank branch)
in the past 12 months, and if so, how many times.32 Since
2013, the survey has measured the share of households that
accessed their accounts using bank tellers. However, some
households may rely on bank branches for activities other
than accessing an account, such as resolving a problem or
asking about products or services, and the questions on account access methods provide only an imprecise measure of
the intensity of branch use. The goal of the new questions is
to provide a more complete picture of household use of bank
branches.
Figure 4.1 shows the distribution of bank branch visits among
banked households.33 Overall, 86.0 percent of banked households visited a bank branch in the past 12 months. About one
in three (30.8 percent) banked households visited a branch
one to four times, 18.2 percent visited five to nine times, and
35.4 percent visited ten or more times. The remaining 14.0
percent of banked households did not visit a branch in the
past 12 months.34
Households that spoke with a teller or other employee in person at a bank branch were asked whether they did so one to four times in the past 12 months, five to nine
times in the past 12 months, or ten or more times in the past 12 months.
32
The analysis of bank branch visits for banked households presented in this section excludes 1,048 observations (representing roughly 4.2 million banked households)
with missing information on whether the household visited a bank branch in the past 12 months.
33
Among unbanked households, 14.7 percent visited a bank branch in the past 12 months: 7.7 percent visited a branch one to four times, 2.2 percent visited five to
nine times, and 4.7 percent visited ten or more times. Approximately two-thirds of unbanked households that visited a branch did not have a bank account at any time
in the past 12 months. (See Appendix Table B.13 for detailed estimates of bank branch visits among unbanked households by previous banking status and household
characteristics.)
34
28 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Figure 4.1 Bank Branch Visits in Past 12 Months Among
Banked Households, 2017 (Percent)
35.4
30.8
18.2
14.0
0 times
1 to 4
times
5 to 9
times
10 or more
times
Note: Households that visited a branch but with unknown frequency
(1.6 percent of banked households) are not shown.
Table 4.6 shows bank branch visits among banked
households by the primary method used to access an
account. About two-thirds (67.8 percent) of banked
households that used bank tellers as their primary method
visited a branch ten or more times. Branch visits were
prevalent even among banked households that used online or
mobile banking as their primary method of account access.
For example, 81.0 percent of banked households that used
mobile banking as their primary method visited a branch in
the past 12 months, and nearly one-quarter (23.0 percent)
visited ten or more times.
Patterns of bank branch visits among banked households
varied by household characteristics, as shown in Table 4.7.
For example, older households, households in rural areas,
and households with volatile income were more likely to visit
a branch or to have visited ten or more times. Black, Hispanic,
and Asian households were less likely to visit a branch or to
have visited ten or more times. While lower-income and lesseducated households were less likely to visit a branch overall,
those that did visit a branch were more likely to have visited
ten or more times.35 (See Appendix Table B.12 for bank
branch visits among banked households that visited
a branch.)
Table 4.6 Bank Branch Visits in Past 12 Months Among Banked Households by Selected Primary Methods of Account
Access, 2017
For all banked households, row percent
0 times
(Percent)
1 to 4 times
(Percent)
5 to 9 times
(Percent)
10 or more times
(Percent)
14.0
30.8
18.2
35.4
Bank teller
0.0
16.1
14.1
67.8
Online banking
15.7
35.6
20.5
27.2
Mobile banking
19.0
38.7
18.8
23.0
All
Primary method of account
access
Note: Households that visited a branch but with unknown frequency (1.6 percent of banked households) are not shown.
The finding that lower-income and less-educated households were more likely to have visited a branch ten or more times is consistent with patterns for bank account
access methods: lower-income and less-educated households were more likely to use bank tellers as their primary or only method of account access.
35
29
Table 4.7 Bank Branch Visits in Past 12 Months Among Banked Households by Banking Status and Selected Household
Characteristics, 2017
For all banked households, row percent
0 times
(Percent)
1 to 4 times
(Percent)
5 to 9 times
(Percent)
10 or more times
(Percent)
14.0
30.8
18.2
35.4
Underbanked
12.6
32.4
17.4
37.0
Fully banked
14.3
30.7
18.6
35.5
Less than $15,000
20.0
28.9
13.5
35.7
$15,000 to $30,000
14.5
29.1
15.6
38.6
$30,000 to $50,000
13.3
30.3
18.0
36.5
$50,000 to $75,000
12.8
30.7
18.6
36.6
At least $75,000
13.1
32.3
20.2
32.9
No high school diploma
17.8
27.1
13.2
39.7
High school diploma
13.6
29.3
15.9
39.3
Some college
13.2
30.1
18.7
36.5
College degree
14.0
33.1
20.3
31.2
15 to 24 years
16.9
34.7
17.5
30.1
25 to 34 years
19.0
35.6
18.0
26.1
35 to 44 years
16.7
35.2
17.9
28.1
45 to 54 years
12.8
30.9
19.1
35.7
55 to 64 years
11.8
28.7
18.0
40.0
65 years or more
10.9
25.7
18.0
43.5
Black
18.0
35.3
17.7
27.2
Hispanic
17.2
35.3
16.0
29.7
Asian
19.6
34.1
17.3
26.7
White
12.2
29.0
18.7
38.4
Other
13.2
27.0
15.6
43.1
Disabled, age 25 to 64
15.6
30.2
16.1
37.1
Not disabled, age 25 to 64
14.8
32.7
18.5
32.3
Income was about the same each month
14.7
31.4
18.4
34.7
Income varied somewhat from month to month
10.9
31.0
18.5
38.8
Income varied a lot from month to month
9.3
25.7
17.0
46.7
Metropolitan area - principal city
17.1
33.5
17.9
29.8
Metropolitan area - balance
14.0
32.8
18.7
32.7
Not in metropolitan area
9.5
22.5
17.9
49.0
Not identified
11.4
26.8
17.4
42.6
Characteristics
All
Banking status
Family income
Education
Age group
Race/Ethnicity
Disability status
Monthly income volatility
Metropolitan status
Notes: Households that visited a branch but with unknown frequency (1.6 percent of banked households) are not shown. See Appendix Table B.11 for estimates by
other household characteristics.
30 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Mobile Activities Among Banked Households
The 2017 survey included a series of questions about the
ways households used a mobile phone for banking activities
in the past 12 months. Most of these activities were also
asked about in the 2013 survey.36
Figure 4.2 presents use of mobile activities among banked
households in 2013 and 2017.37 Use of a mobile phone to
check email from a bank about an account was the most
common activity in 2017, performed by 44.1 percent of
banked households.38 Other common activities, performed
by about one-third of banked households in 2017, were using
a bank’s mobile website or bank’s mobile app to check a
bank account balance or recent transactions, and receiving
a mobile text alert or push notification from a bank about an
account. Use of a mobile phone to check a bank account balance or recent transactions increased substantially from 19.0
percent in 2013 to 35.4 percent in 2017. The remaining mobile
activities asked about in the survey were less common, but
the proportion of banked households that performed each
of these activities doubled or more than doubled from 2013
to 2017. Growth in the use of a mobile phone’s camera to
deposit a check into a bank account was particularly striking,
as the proportion of banked households that performed this
activity more than tripled, from 5.6 percent in 2013 to 18.0
percent in 2017.
Figure 4.2 Mobile Activities Among Banked Households by Year (Percent)
Checked email about an account
44.1
19.0
Checked balance or transactions
35.4
Text message alert
34.0
13.2
Bill payment
26.5
12.2
Transferred money between accounts
Deposited a check electronically
Sent money to others
25.4
5.6
18.0
5.9
13.7
2013
2017
Note: Estimates of the proportion of banked households that used a mobile phone to check email from a bank about an account or that received a mobile text alert
or push notification from a bank about an account are not available for 2013.
All of the activities asked about in the 2017 survey, except for whether a household used a mobile phone to check email from a bank about an account, were also
asked in the 2013 survey. The 2013 survey included some activities not asked about in the 2017 survey. See Appendix 2 for additional details.
36
Use of a mobile phone to check email from a bank about an account was asked only in 2017, so estimates are not available for 2013. Different types of households in
the 2013 and 2017 surveys were asked whether they received a mobile text alert or push notification from a bank about an account, so estimates of the proportion of
banked households that performed this activity cannot be compared over time.
37
In the 2017 survey, all banked households were asked whether they used a mobile phone to check email from a bank about an account, even if they did not report that
they used mobile banking to access their accounts. The proportion of banked households that checked email from a bank about an account (44.1 percent) was higher
than the proportion of banked households that used mobile banking to access their accounts in 2017 (40.4 percent; see Table 4.2). The Federal Reserve found similar
discrepancies in the 2017 Survey of Household Economics and Decisionmaking. See “Mobile Banking: A Closer Look at Survey Measures,” March 27, 2018 (available at
http://www.federalreserve.gov/econres/notes/feds-notes/mobile-banking-a-closer-look-at-survey-measures-20180327.htm).
38
31
Table 4.8 Mobile Activities Among Banked Households by Banking Status and Selected Household Characteristics, 2017
For all banked households, row percent
Characteristics
All
Checked
email about
an account
(Percent)
Checked
balance or
transactions
(Percent)
Text message
alert
(Percent)
44.1
35.4
34.0
Bill payment
(Percent)
Transferred
money
between
accounts
(Percent)
Deposited
a check
electronically
(Percent)
Sent money
to others
(Percent)
26.5
25.4
18.0
13.7
Banking status
Underbanked
52.8
42.5
42.3
30.7
29.5
19.7
16.9
Fully banked
44.3
35.7
33.9
27.0
26.1
18.8
13.7
Family income
Less than $15,000
26.4
18.3
18.5
11.1
10.0
6.4
5.8
$15,000 to $30,000
29.6
21.2
20.3
13.8
11.6
7.7
6.0
$30,000 to $50,000
38.8
30.4
28.4
21.6
19.7
12.9
9.9
$50,000 to $75,000
46.0
36.1
34.7
27.4
25.9
17.7
13.0
At least $75,000
56.7
48.1
46.4
37.9
37.8
28.3
21.4
Education
No high school diploma
25.0
14.8
17.5
10.0
8.0
5.0
4.4
High school diploma
33.6
25.2
24.8
17.8
15.6
9.2
7.3
Some college
45.3
36.5
34.1
26.6
25.6
16.2
12.6
College degree
54.3
45.8
43.6
35.6
35.3
28.0
20.7
Age group
15 to 24 years
67.6
59.2
45.2
40.1
39.3
27.3
22.6
25 to 34 years
67.6
60.7
50.6
45.0
44.4
34.4
26.5
35 to 44 years
59.8
50.2
45.2
38.2
36.6
26.9
19.3
45 to 54 years
49.0
39.6
39.4
30.3
28.8
19.2
14.6
55 to 64 years
37.0
25.6
30.5
19.1
18.0
11.6
9.4
65 years or more
17.0
10.2
13.4
7.7
6.9
4.4
3.1
46.3
35.0
38.9
25.3
24.1
14.8
15.4
Race/Ethnicity
Black
Hispanic
49.6
37.0
39.8
29.5
25.0
16.2
15.1
Asian
53.5
41.7
44.4
33.4
28.4
25.3
21.5
White
42.0
34.6
31.3
25.5
25.3
18.3
12.5
Other
45.9
39.3
35.3
29.5
29.7
16.7
13.5
Disabled, age 25 to 64
32.3
24.0
25.2
17.1
15.1
9.8
7.6
Not disabled, age 25 to 64
55.0
45.5
42.9
34.4
33.3
24.0
18.2
Income was about the same
each month
44.3
35.5
34.0
26.8
25.5
18.1
13.9
Income varied somewhat
from month to month
53.8
44.9
42.1
32.8
32.3
22.7
16.5
Income varied a lot from
month to month
51.9
43.6
41.3
30.0
32.4
24.9
18.3
Disability status
Monthly income volatility
Note: See Appendix Table B.14 for estimates by other household characteristics.
32 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Table 4.8 shows use of mobile activities among banked
households in 2017 by banking status and selected household characteristics. Underbanked households were more
likely to perform each activity than fully banked households.
The three largest differences between underbanked and fully
banked households were in checking email from a bank about
an account, checking a bank account balance or recent transactions, and receiving a mobile text alert or push notification
from a bank about an account. These results are consistent
with the types of activities described in focus groups with underserved consumers conducted by the FDIC in 2015. Some
consumers who used mobile financial services reported that
mobile alerts and monitoring tools helped them reduce fees,
better track their finances, and improve on-the-spot
decision making.39
Use of each mobile activity was more common among
higher-income households, more-educated households,
younger households, working-age nondisabled households,
and households with volatile income. In many cases, differences in use were substantial across households of different
characteristics. For example, 45.0 percent of households
aged 25 to 34 used a bank’s mobile website or bank’s mobile
app to make a bill payment, compared with only 7.7 percent
of households aged 65 or older. Some differences were also
observed in the use of mobile activities by race and ethnicity.
For example, black and Hispanic households were more likely
than white households to receive a mobile text alert or push
notification from a bank about an account and less likely to
use a mobile phone’s camera to deposit a check into a bank
account. These differences, however, were generally smaller
than differences by income, education, age, and disability
status.
See “Opportunities for Mobile Financial Services to Engage Underserved Consumers Qualitative Research Findings,” May 25, 2016 (available at http://www.fdic.gov/
consumers/community/mobile/MFS_Qualitative_Research_Report.pdf).
39
33
2017 FDIC National Survey of Unbanked and Underbanked
Households
5. Prepaid Cards
Some consumers use general purpose reloadable prepaid
cards to address their financial transactions needs. Similar
to a checking account, these cards can be used to pay bills,
withdraw cash at ATMs, make purchases, deposit checks,
and receive direct deposits. Consumers can obtain prepaid
cards from sources such as a bank location or bank’s website, a nonbank store or website, a government agency, or an
employer.40 Many, although not all, such cards store funds in
accounts eligible for deposit insurance.
As in the 2013 and 2015 surveys, the 2017 survey asked
households whether they used a prepaid card in the past
12 months, referred to in this report as prepaid card use.41
Between 2015 and 2017, the proportion of households
that used prepaid cards decreased from 9.8 percent to 9.2
percent, as shown in Table 5.1. This decline can be attributed
primarily to changes in income and other characteristics
of U.S. households between 2015 and 2017. However, the
proportion of households that used prepaid cards in 2017
remained higher than in 2013 (7.9 percent).42
Table 5.1 Prepaid Card Use in Past 12 Months by Year
For all households, row percent
Year
Number of
Households
(1000s)
Used
(Percent)
Did not use
(Percent)
Unknown
(Percent)
2013
123,750
7.9
86.4
5.7
2015
127,538
9.8
85.8
4.4
2017
129,276
9.2
85.5
5.4
Prepaid Card Use by Household Characteristics
Differences in prepaid card use across households were
similar in 2017 to earlier years. As shown in Table 5.2, prepaid card use was higher among lower-income households,
less-educated households, younger households, black
households, working-age disabled households, and households with volatile income.43
Mirroring the national trend, prepaid card use among many
socioeconomic and demographic groups declined from 2015
to 2017, though the declines were often not statistically significant. One exception is households aged 15 to 24, where
prepaid card use increased from 12.4 percent in 2015 to 15.1
percent in 2017. For most socioeconomic and demographic
groups, prepaid card use in 2017 remained higher than in
2013.
Prepaid Card Use by Geography
As in previous years, prepaid card use varied across regions
of the United States. In 2017, 7.9 percent of households in
the Northeast used prepaid cards, compared with 8.7 percent
in the West, 9.5 percent in the South, and 10.0 percent in
the Midwest. Figure 5.1 shows that prepaid card use varied
considerably across states in 2017, ranging from 4.6 percent
in Hawaii to 15.4 percent in Mississippi. (See Appendix Tables
C.3 and C.4 for detailed state- and MSA-level estimates and
for selected confidence intervals.)
40
Households were instructed that the survey questions about prepaid cards were “not asking about gift cards or debit cards linked to a checking account.”
41
For the 2017 survey, a small change was made to the introductory description of prepaid cards. See Appendix 2 for details.
After accounting for changes in bank account ownership and the household characteristics listed in Appendix Table A.2, the remaining decrease in prepaid card use
from 2015 to 2017 was not statistically significant, while the remaining increase in prepaid card use from 2013 to 2017 was statistically significant.
42
Some of these differences in prepaid card use, namely by income level, education, and race, can be explained by differences in bank account ownership and other
household characteristics listed in Appendix Table A.2.
43
34 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Table 5.2 Prepaid Card Use in Past 12 Months by Selected Household Characteristics and Year
For all households
2013
(Percent)
2015
(Percent)
2017
(Percent)
Difference
(2017-2015)
7.9
9.8
9.2
-0.6*
Less than $15,000
11.4
14.3
13.0
-1.3
$15,000 to $30,000
8.3
10.8
10.4
-0.4
$30,000 to $50,000
8.3
8.8
9.1
0.3
$50,000 to $75,000
6.4
9.2
7.7
-1.5*
At least $75,000
6.5
8.1
8.0
0.0
No high school diploma
8.9
11.0
10.3
-0.7
High school diploma
8.1
10.3
9.3
-1.0*
Some college
8.8
10.8
10.0
-0.8
College degree
6.7
8.0
8.0
0.0
15 to 24 years
12.7
12.4
15.1
2.7*
25 to 34 years
10.9
12.6
11.0
-1.6*
35 to 44 years
10.3
11.4
10.8
-0.6
45 to 54 years
9.1
11.0
10.7
-0.2
55 to 64 years
6.4
9.3
8.7
-0.6
65 years or more
3.0
5.5
4.9
-0.6
Black
11.5
13.9
13.3
-0.5
Hispanic
7.8
9.6
7.9
-1.7*
Asian
4.4
5.7
7.2
1.5
White
7.3
9.1
8.5
-0.6*
Other
13.8
17.0
15.9
-1.1
Disabled, age 25 to 64
12.4
15.2
15.7
0.5
Not disabled, age 25 to 64
8.7
10.4
9.4
-0.9*
Income was about the same each
month
9.2
8.9
-0.3
Income varied somewhat from month
to month
13.5
12.6
-1.0
Income varied a lot from month to
month
15.5
12.8
-2.7*
Characteristics
All
Family income
Education
Age group
Race/Ethnicity
Disability status
Monthly income volatility
Notes: Monthly income volatility is not available for 2013. Asterisks indicate differences that are statistically significant at the 10 percent level. See Appendix Table
C.1 for estimates by other household characteristics and for selected confidence intervals.
35
Figure 5.1 Prepaid Card Use in Past 12 Months by State, 2017 (Percent)
WA
ND
ND
MT
OR
MN
ID
MI
WY
IL
UT
CO
KS
CA
OK
NM
AZ
MO
WV
KY
TN
AL
LA
MD
VA
MA
CT
NJ
RI
DE
DC
NC
AR
MS
TX
PA
OH
IN
VT
NH
NY
IA
NE
NV
ME
WI
SD
SC
GA
Less than 7.8
7.8 to 9.0
FL
AK
HI
9.0 to 9.7
9.7 to 11.1
At least 11.1
Prepaid Card Use by Banking Status
Use of prepaid cards in 2017 was most prevalent among
unbanked households, consistent with previous survey
results. As illustrated in Figure 5.2, 26.9 percent of unbanked
households used a prepaid card in 2017, compared with 14.5
percent of underbanked households and 6.7 percent of fully
banked households. These percentages are similar to 2015. In
line with the increase in prepaid card use from 2013 to 2017
at the national level, prepaid card use within each banking
status group increased over this period.
Figure 5.2 Prepaid Card Use in Past 12 Months by Banking
Status and Year (Percent)
Overall, approximately half (48.7 percent) of households that
used prepaid cards in 2017 were either unbanked or underbanked, as shown in Figure 5.3. For context, this percentage
is almost double the 25.2 percent of all households that were
either unbanked or underbanked in 2017 (see Figure 3.1).
Figure 5.3 Banking Status of Households That Used
Prepaid Cards in Past 12 Months, 2017 (Percent)
Banked,
underbanked
status unknown
1.3
Underbanked
29.6
27.1 26.9
22.3
13.1
15.4 14.5
5.3
Unbanked
Underbanked
2013
2015
6.9
Fully
banked
50.0
6.7
Fully banked
2017
Unbanked
19.1
36 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Figure 5.4 Cited Reasons for Not Having a Bank Account by Prepaid Card Use, 2017 (Percent)
53.1
Do not have enough money to keep in account
Don't trust banks
30.2
Account fees too high
29.6
29.9
Account fees unpredictable
18.8
ID, credit, or former bank account problems
16.4
14.3
Banks do not offer needed products or services
15.9
13.2
Inconvenient locations
Other reason
Though many of the cited reasons for not having a bank
account were similar for households that used prepaid cards
and those that did not, in some cases differences existed
between these groups. For example, as illustrated in Figure
5.4, unbanked households that used prepaid cards were
more likely than those that did not use prepaid cards to cite
the following reasons for not having an account: “Don’t trust
banks,” “Bank account fees are too high,” and “Bank account
fees are unpredictable.” Irrespective of prepaid card use, the
most commonly cited reason for not having a bank account
was “Do not have enough money to keep in an account.”44
Sources of Prepaid Cards
Consistent with previous survey results, households that used
prepaid cards in 2017 obtained them from a variety of sourc-
44
27.1
10.0
9.8
9.0
9.8
16.5
15.6
Used prepaid card
Unbanked households that used prepaid cards were more
likely to have had a bank account at some point in the past:
62.7 percent of unbanked households that used prepaid
cards in 2017 had a bank account in the past, compared
with 41.9 percent of unbanked households that did not use
prepaid cards.
34.3
22.3
Avoiding banks gives more privacy
Inconvenient hours
57.8
37.0
Did not use prepaid card
es. As shown in Figure 5.5, the most common source in 2017
was a store or website that is not a bank (45.4 percent of
households that used prepaid cards obtained cards from this
source), followed by a government agency, family or friends,
and a bank location or a bank’s website.
Among households that used prepaid cards, the proportion
that obtained cards from stores or websites that are not
banks increased from 42.6 percent in 2015 to 45.4 percent in
2017, while the proportion that obtained cards from banks decreased from 17.3 percent to 13.3 percent.45 The decline from
2015 to 2017 in the use of prepaid cards obtained from banks
accounts for most of the decline in overall prepaid card use
during this period.46
As in 2015, sources of prepaid cards differed by banking
status in 2017 (see Table 5.3). In particular, unbanked households that used prepaid cards were much less likely to have
obtained cards from banks, compared with underbanked and
fully banked households that used prepaid cards. Regardless
of banking status, the most common source of prepaid cards
in 2017 was a store or website that is not a bank.
The main reasons for not having a bank account showed qualitatively similar patterns as the cited reasons in Figure 5.4.
Estimates of the share of households that obtained prepaid cards from the various sources in 2015 and 2017 were not comparable to 2013 because of changes to
the survey instrument.
45
Further, after accounting for changes in bank account ownership and the household characteristics listed in Appendix Table A.2, the remaining decrease from 2015 to
2017 in the use of prepaid cards from banks was statistically significant, unlike for overall prepaid card use.
46
37
Figure 5.5 Sources of Prepaid Cards for Households That Used Prepaid Cards in Past 12 Months by Year (Percent)
42.6
Store or website that is not a bank
45.4
Government agency
14.8
15.0
Family or friends
14.2
15.0
17.3
Bank location or bank's website
13.3
9.2
9.3
Employer payroll card
6.8
8.4
Other
Unknown
1.3
0.6
2015
2017
Note: Bars sum to more than 100 percent because households with multiple prepaid cards were asked to select all sources of their cards.
Table 5.3 Sources of Prepaid Cards by Banking Status, 2017
For all households that used prepaid cards in the past 12 months, row percent
Store or
website that
is not a bank
(Percent)
Government
agency
(Percent)
Family or
friends
(Percent)
Bank
location
or bank's
website
(Percent)
Employer
payroll card
(Percent)
Other
(Percent)
Unknown
(Percent)
45.4
15.0
15.0
13.3
9.3
8.4
0.6
Unbanked
44.6
31.5
3.1
5.1
13.9
7.6
0.6
Underbanked
48.4
12.4
13.5
14.9
9.6
8.3
0.6
Fully banked
44.2
10.3
19.7
15.7
7.3
8.9
0.6
Characteristics
All
Banking status
Note: Row percentages sum to more than 100 because households with multiple prepaid cards were asked to select all sources of their cards.
38 | 2017 FDIC National Survey of Unbanked and Underbanked Households
2017 FDIC National Survey of Unbanked and Underbanked
Households
6. Alternative Financial Services
Alternative Financial Services Use by Household
Characteristics
As in earlier surveys, the 2017 survey asked households
about their use of alternative financial services (AFS) during
the past 12 months. Households were asked if they went to
a place other than a bank to purchase a money order, cash a
check, or send an international remittance (transaction AFS).
Households were also asked whether they used any of the
following nonbank products and services that may be used
in lieu of bank credit: payday loans, refund anticipation loans,
rent-to-own services, pawn shop loans, and auto title loans
(credit AFS).47
Consistent with past survey results, AFS use differed across
households. As shown in Table 6.2, AFS use in 2017 was
more common among lower-income households, less-educated households, younger households, black and Hispanic
households, working-age disabled households, and households with volatile income.
Declines in AFS use over time were fairly widespread across
segments of the population. For example, among households
with less than $15,000 in income, 32.4 percent used AFS in
2017, down from 38.6 percent in 2015 and 39.1 percent in
2013. AFS use among younger households, black households, and working-age disabled households also decreased
substantially from 2013 to 2017.
As shown in Table 6.1, 22.1 percent of households used
some type of AFS in 2017, down from 24.0 percent in 2015
and 24.9 percent in 2013. Use of transaction AFS remained
more common than use of credit AFS in 2017. However, use
of transaction AFS steadily declined from 2013, while use of
credit AFS decreased from 2015 to 2017 after having increased from 2013 to 2015. The decrease in the use of credit
AFS from 2015 to 2017 can be attributed primarily to changes
in income and other characteristics of U.S. households over
this period.48
Table 6.1 Alternative Financial Services Use in Past 12 Months by Year
For all households, row percent
Any AFS
Transaction AFS
Credit AFS
Year
Number of
Households
(1000s)
Used
(Percent)
Did not use
(Percent)
Unknown
(Percent)
Used
(Percent)
Did not use
(Percent)
Unknown
(Percent)
Used
(Percent)
Did not use
(Percent)
Unknown
(Percent)
2013
123,750
24.9
69.3
5.8
21.9
72.9
5.2
7.0
87.2
5.8
2015
127,538
24.0
70.3
5.8
20.2
74.2
5.6
7.7
86.9
5.5
2017
129,276
22.1
70.9
7.1
18.3
74.8
6.9
6.9
86.4
6.7
In this section, all estimates of AFS use are based on the 12 months before the survey. The 2017 survey asked about the same set of AFS as the 2013 and 2015
surveys, except that the 2017 survey included a new question that asked households that had not used payday loans, refund anticipation loans, pawn shop loans, or
auto title loans whether they had, in the past 12 months, “taken out any other types of loans or lines of credit from a payday lender, auto title lender, pawn shop, or
check casher.” This question was designed to elicit use of installment loans from payday lenders, auto title lenders, pawn shops, or check cashers. Because only 0.5
percent of households that were asked this question responded affirmatively, the 2017 estimates of AFS use and banking status in this report do not incorporate the
responses to this question and are therefore comparable to the 2013 and 2015 estimates. Incorporating the responses to this question would lead to small changes
in estimated credit AFS use, any AFS use, and the underbanked rate for 2017: the proportion of households that used credit AFS would increase from 6.9 percent to
7.3 percent, the proportion of households that used any AFS would increase from 22.1 percent to 22.4 percent, and the underbanked rate would increase from 18.7
percent to 19.0 percent. See Appendix 2 for additional details on changes to the survey instrument.
47
After accounting for changes in the household characteristics listed in Appendix Table A.2, the remaining decrease in the use of credit AFS from 2015 to 2017 was
not statistically significant.
48
39
Table 6.2 Alternative Financial Services Use by Selected Household Characteristics and Year
For all households
2013
(Percent)
2015
(Percent)
2017
(Percent)
Difference
(2017-2015)
24.9
24.0
22.1
-1.9*
Less than $15,000
39.1
38.6
32.4
-6.2*
$15,000 to $30,000
33.1
31.0
29.7
-1.2
Characteristics
All
Family income
$30,000 to $50,000
26.5
26.6
25.8
-0.9
$50,000 to $75,000
20.9
21.0
20.5
-0.5
At least $75,000
13.8
13.5
13.5
-0.1
Education
No high school diploma
39.5
39.0
35.4
-3.6*
High school diploma
28.7
27.8
25.3
-2.5*
Some college
26.9
25.3
23.6
-1.8*
College degree
14.9
15.0
14.9
-0.1
Age group
15 to 24 years
41.5
37.8
34.6
-3.2
25 to 34 years
33.6
31.1
28.0
-3.1*
35 to 44 years
29.6
28.1
26.6
-1.5
45 to 54 years
26.7
24.8
22.9
-1.9*
55 to 64 years
20.9
21.8
20.7
-1.1
65 years or more
13.1
14.1
13.0
-1.1*
Race/Ethnicity
Black
46.1
42.2
39.9
-2.3*
Hispanic
40.3
38.5
36.0
-2.4*
Asian
18.6
22.3
18.0
-4.3*
White
18.1
17.3
15.5
-1.8*
Other
36.5
34.1
35.0
0.9
Disabled, age 25 to 64
38.7
38.0
33.2
-4.8*
Not disabled, age 25 to 64
26.0
24.5
23.1
-1.4*
22.6
21.2
-1.4*
Disability status
Monthly income volatility
Income was about the same each month
Income varied somewhat from month to month
32.6
30.6
-1.9*
Income varied a lot from month to month
40.3
36.4
-3.9*
Notes: Monthly income volatility is not available for 2013. Asterisks indicate differences that are statistically significant at the 10 percent level. See Appendix Table
D.1 for estimates by other household characteristics and for selected confidence intervals.
40 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Table 6.3 Use of Specific Alternative Financial Services by Bank Account Ownership and Year
For all households
2013
(Percent)
2015
(Percent)
2017
(Percent)
Difference
(2017-2015)
Money orders
17.3
15.0
13.4
-1.6*
Check cashing
6.5
6.5
5.9
-0.6*
Remittances
3.7
3.7
3.4
-0.3
Pawn shop loans
2.9
1.8
1.4
-0.4*
Payday loans
2.0
2.0
1.7
-0.3*
Refund anticipation loans
1.8
2.6
2.4
-0.2*
Rent-to-own
1.5
1.8
1.4
-0.3*
Auto title loans
0.9
1.3
1.4
0.0
Money orders
47.3
43.2
39.5
-3.7*
Check cashing
35.9
30.3
27.0
-3.3*
Remittances
9.2
7.9
5.6
-2.3*
Pawn shop loans
9.9
6.6
4.3
-2.3*
Payday loans
2.7
3.6
2.9
-0.7
Refund anticipation loans
3.8
4.5
3.3
-1.1
Rent-to-own
4.5
5.0
3.6
-1.4*
Auto title loans
1.7
2.3
2.2
-0.1
Specific AFS
A. All households
B. Unbanked households
C. Banked households
Money orders
14.7
12.8
11.6
-1.2*
Check cashing
4.1
4.7
4.4
-0.2
Remittances
3.2
3.4
3.3
-0.1
Pawn shop loans
2.3
1.5
1.2
-0.2*
Payday loans
1.9
1.8
1.6
-0.2*
Refund anticipation loans
1.6
2.5
2.3
-0.2
Rent-to-own
1.2
1.5
1.3
-0.2*
Auto title loans
0.8
1.3
1.3
0.1
Note: Asterisks indicate differences that are statistically significant at the 10 percent level.
41
Alternative Financial Services Use by Bank Account
Ownership
AFS use continued to be much higher among unbanked
households than banked households. As shown in Figures
6.1 and 6.2, 51.3 percent of unbanked households used AFS
in 2017, compared with 20.0 percent of banked households.
Although AFS use was more common among unbanked
households, the proportion that used AFS decreased substantially in recent years and is attributable to declines in the
use of both transaction and credit AFS over this period. Use
of AFS among banked households also decreased in recent
years and is attributable almost entirely to the decline in the
use of transaction AFS over this period.
Table 6.3 shows use of specific AFS by bank account ownership and year. Overall, as shown in panel A, money orders
remained the most commonly used AFS in 2017, followed by
check cashing and remittances.49 However, use of nonbank
money orders declined substantially in recent years,
particularly among unbanked households. Use of nonbank
check cashing also declined sharply among unbanked households: 27.0 percent used a nonbank check casher in 2017,
down from 30.3 percent in 2015 and 35.9 percent in 2013.
Figure 6.1 Alternative Financial Services Use in Past 12
Months by Year, Unbanked Households (Percent)
63.2
57.3
60.5
51.3
54.1
48.0
16.7 16.5
Any AFS
Transaction AFS
2013
2015
12.3
Credit AFS
2017
Figure 6.2 Alternative Financial Services Use in Past 12
Months by Year, Banked Households (Percent)
21.7 21.4 20.0
18.6 17.6 16.3
6.2
Any AFS
Transaction AFS
2013
2015
7.0
6.5
Credit AFS
2017
In addition to nonbank remittance use, the 2015 and 2017 surveys asked households whether they used a bank to send a remittance in the past 12 months. Similar
proportions of households used banks to send remittances in 2015 (1.6 percent) and 2017 (1.8 percent). In both years, use of banks to send remittances was less
common than use of nonbanks.
49
42 | 2017 FDIC National Survey of Unbanked and Underbanked Households
2017 FDIC National Survey of Unbanked and Underbanked
Households
7. Saving for Unexpected Expenses or Emergencies
Savings can help households better manage unexpected expenses or emergencies, such as a sudden illness, job loss, or
home or car repairs. The absence of savings can sometimes
be a barrier to financial stability and resilience, particularly for
consumers with uneven or low incomes. To gain insight into
these issues, the 2015 and 2017 surveys included questions
on whether households saved for unexpected expenses or
emergencies and the methods they used.
Mirroring the increase in the savings rate from 2015 to 2017
at the national level, savings rates for many population segments increased over the same period, though the increases
were often not statistically significant. The savings rate among
Hispanic households increased considerably, from 42.5 percent in 2015 to 48.2 percent in 2017. Moreover, savings rates
among younger households increased more than savings
rates among older households.
Specifically, households were asked whether they set aside
any money in the past 12 months that could be used for
unexpected expenses or emergencies, even if the funds were
later spent.50 Households were prompted to consider only
Savings Rates by Geography
funds that could have been easily spent, if necessary, and
not retirement or other long-term savings. Households that
set aside money for this purpose were then asked where
they kept the money, indicating one or more of the following
methods: savings accounts; checking accounts; prepaid
cards; other accounts such as certificates of deposit, brokerage accounts, or savings bonds; in the home, or with family
or friends; buying something with the intent to pawn it or sell
it later, if necessary; or other methods.
Between 2015 and 2017, the proportion of households that
saved for unexpected expenses or emergencies in the past
12 months increased from 56.3 percent to 57.8 percent.51
This increase can be attributed primarily to changes in income
and other characteristics of U.S. households between 2015
and 2017.52
Savings Rates by Household Characteristics
As in 2015, rates of saving for unexpected expenses or emergencies in 2017 varied by household characteristics (see Table
7.1). For example, savings rates were lower among lower-income households, less-educated households, older households,
black and Hispanic households, and working-age disabled
households. Differences by income and education were especially pronounced. For instance, only 28.9 percent of households
with less than $15,000 in income saved for unexpected expenses or emergencies in 2017, compared with 73.8 percent of
households with income of $75,000 or more.
As in 2015, rates of saving for unexpected expenses or
emergencies in 2017 varied across regions of the United
States. Savings rates continued to be higher in the Midwest
(60.9 percent in 2017) and West (60.4 percent) and lower in
the Northeast (55.8 percent) and South (55.3 percent). The
savings rate in the South increased from 52.1 percent in 2015
to 55.3 percent in 2017, the largest increase among regions.
(See Appendix Table E.1.)
Figure 7.1 shows that savings rates varied widely across
states in 2017, ranging from 40.5 percent in West Virginia
to 74.1 percent in Utah. Certain areas experienced large
changes in savings rates from 2015 to 2017. For example, the
savings rate in Washington, DC, increased from 53.2 percent
to 67.0 percent, while the savings rate in Maine decreased
from 68.7 percent to 60.0 percent. (See Appendix Tables E.2
and E.3 for detailed state- and MSA-level estimates and for
selected confidence intervals.)
Savings Methods
Figure 7.2 shows that among all households that saved for
unexpected expenses or emergencies, savings and checking
accounts were the most used savings methods in 2015 and
2017: more than four in five (85.5 percent in 2017 and 84.9
percent in 2015) kept savings in one of these accounts. In
both years, about one in ten (10.5 percent) households that
saved maintained savings in the home, or with family
or friends.
The question allows for funds to be later spent because a household might have experienced an unexpected expense or emergency that required the household to
draw on the money that had been saved.
50
The analysis presented in this section excludes 3,106 observations (representing roughly 12.1 million households) with missing information on whether the household
saved for unexpected expenses or emergencies.
51
After accounting for changes in the household characteristics listed in Appendix Table A.2, the remaining increase in the savings rate from 2015 to 2017 was not
statistically significant.
52
43
Table 7.1 Rates of Saving for Unexpected Expenses or
Emergencies by Selected Household Characteristics
and Year
For all households
2015
(Percent)
2017
(Percent)
Difference
(2017-2015)
56.3
57.8
1.4*
Less than $15,000
30.8
28.9
-1.9
Characteristics
All
Family income
$15,000 to $30,000
42.2
41.0
-1.2
$30,000 to $50,000
53.2
54.7
1.4
$50,000 to $75,000
63.6
63.7
0.1
At least $75,000
72.9
73.8
1.0
No high school diploma
30.1
31.7
1.6
High school diploma
47.2
48.6
1.4
Some college
58.9
59.1
0.2
College degree
69.4
70.0
0.6
15 to 24 years
55.7
60.1
4.4*
25 to 34 years
60.7
63.9
3.1*
35 to 44 years
58.8
61.2
2.4*
45 to 54 years
58.2
59.3
1.1
55 to 64 years
56.4
56.8
0.4
65 years or more
50.1
50.6
0.5
Black
45.6
45.7
0.1
Hispanic
42.5
48.2
5.7*
Asian
52.9
55.2
2.3
White
61.3
62.4
1.1*
Other
56.2
53.1
-3.1
Disabled, age 25 to 64
39.0
38.5
-0.5
Not disabled, age 25
to 64
61.3
63.2
1.9*
Income was about the
same each month
56.3
56.9
0.7
Income varied somewhat
from month to month
58.3
63.0
4.7*
Income varied a lot from
month to month
51.3
52.5
1.2
Education
Savings Methods by Household Characteristics
As in 2015, the use of formal (e.g., savings or checking
accounts) and informal (e.g., in the home, or with family or
friends) savings methods varied by household characteristics
in 2017 (see Table 7.2). For example, among households
that saved for unexpected expenses or emergencies, lowerincome households, less-educated households, working-age
disabled households, and households with volatile income
were less likely to keep savings in a savings account and
more likely to maintain savings in the home, or with family
or friends. As with savings rates, differences in savings
methods by income and education were considerable. For
instance, among households that saved, 49.6 percent of
households with less than $15,000 in income kept savings
in a savings account in 2017, compared with 79.1 percent of
households with income of $75,000 or more.53 Differences in
savings methods by race and ethnicity were small relative to
differences by income and education.
Age group
Race/Ethnicity
Disability status
Monthly income
volatility
Notes: Asterisks indicate differences that are statistically significant at the
10 percent level. See Appendix Table E.1 for estimates by other household
characteristics and for selected confidence intervals.
Among households that saved, 71.0 percent of households with less than $15,000 in income kept savings in a savings or checking account in 2017, compared with
89.9 percent of households with income of $75,000 or more.
53
44 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Figure 7.1 Rates of Saving for Unexpected Expenses or Emergencies by State, 2017
WA
ND
MT
OR
MN
ID
MI
WY
UT
CO
KS
CA
AZ
OK
NM
MO
WV
KY
TN
AL
LA
MD
VA
MA
CT
NJ
RI
DE
DC
NC
AR
MS
TX
PA
OH
IN
IL
VT
NH
NY
IA
NE
NV
ME
WI
SD
SC
GA
Less than 55.4
55.4 to 58.8
FL
AK
HI
58.8 to 60.2
60.2 to 66.3
At least 66.3
Figure 7.2 All Savings Methods for Households That Saved by Year (Percent)
70.1
71.6
Savings account
24.5
23.7
Checking account
10.5
10.5
In home, or with family or friends
9.5
9.0
Other accounts
Prepaid card
0.5
0.3
Intent to pawn or sell
0.2
0.1
Other method
Unknown method
2.0
2.6
1.4
1.5
2015
2017
Note: Bars sum to more than 100 percent because households were asked to select all savings methods used.
45
Table 7.2 Savings Methods by Selected Household Characteristics, 2017
For all households that saved for unexpected expenses or emergencies in the past 12 months, row percent
Characteristics
All
Savings account
(Percent)
Checking account
(Percent)
In home, or with
family or friends
(Percent)
Prepaid card
(Percent)
71.6
23.7
10.5
0.3
Family income
Less than $15,000
49.6
27.7
23.7
1.1
$15,000 to $30,000
54.3
26.2
19.1
0.7
$30,000 to $50,000
67.2
24.0
13.2
0.6
$50,000 to $75,000
74.9
21.8
9.2
0.2
At least $75,000
79.1
23.3
6.1
0.1
52.8
26.0
21.6
1.1
Education
No high school diploma
High school diploma
64.8
23.5
15.5
0.5
Some college
70.1
21.9
12.1
0.4
College degree
78.1
24.8
5.7
0.1
67.7
18.9
16.5
1.0
Age group
15 to 24 years
25 to 34 years
73.2
21.5
11.4
0.6
35 to 44 years
74.3
21.3
11.0
0.2
45 to 54 years
72.8
23.5
10.0
0.4
55 to 64 years
71.8
24.8
10.8
0.1
65 years or more
67.7
27.9
7.9
0.2
67.1
21.5
14.2
1.0
Race/Ethnicity
Black
Hispanic
65.9
23.0
15.5
0.5
Asian
75.4
31.7
3.0
0.1
White
73.0
23.5
9.5
0.2
Other
62.8
27.6
20.5
1.2
Disability status
Disabled, age 25 to 64
58.7
25.9
18.4
0.9
Not disabled, age 25 to 64
74.2
22.5
10.2
0.3
Income was about the same
each month
72.9
23.7
9.4
0.3
Income varied somewhat from
month to month
69.1
23.8
13.5
0.5
Income varied a lot from
month to month
60.2
24.4
18.4
0.6
Monthly income volatility
Notes: Row percentages sum to more than 100 because households were asked to select all savings methods used. See Appendix Table E.4 for the full set of
savings methods and for estimates by other household characteristics.
46 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Savings Rates and Methods by Banking Status
As in 2015, unbanked households saved for unexpected
expenses or emergencies at a much lower rate than underbanked and fully banked households. Figure 7.3 shows that
17.4 percent of unbanked households saved for this purpose
in 2017, compared with 56.3 percent of underbanked households and 61.6 percent of fully banked households. These
estimates are similar to 2015.54
Figure 7.3 Rates of Saving for Unexpected Expenses or
Emergencies by Banking Status and Year
55.2
20.2
56.3
60.0
61.6
17.4
Unbanked
Underbanked
2015
Fully banked
Unbanked households generally saved using informal methods, while underbanked and fully banked households generally saved using formal methods, consistent with results from
the 2015 survey. Figure 7.4 shows that, in 2017, unbanked
households that saved primarily kept savings in the home,
or with family or friends (66.8 percent).55 In contrast, underbanked and fully banked households that saved primarily
used savings accounts. The vast majority of underbanked
(80.2 percent) and fully banked (88.8 percent) households that
saved kept savings in a savings or checking account.
Although prepaid cards are generally thought of as transactional, some households, particularly the unbanked, used
them to save. Among households that saved and used a
prepaid card in the past 12 months, 23.8 percent of unbanked
households kept savings on a prepaid card, compared with
only 2.8 percent of underbanked households and 0.9 percent
of fully banked households.
2017
Figure 7.4 Selected Savings Methods for Households That Saved by Banking Status, 2017 (Percent)
75.0
66.8
65.1
24.1
23.9
16.3
10.1
2.0
7.7
2.2
0.4
Unbanked
Savings account
Underbanked
Checking account
In home, or with family or friends
0.1
Fully banked
Prepaid card
Notes: Bars may sum to more than 100 percent because households were asked to select all savings methods used. See Appendix Table E.4 for the full set of
savings methods by banking status.
The decline in the savings rate for unbanked households from 2015 to 2017 was no longer statistically significant after accounting for changes in the household
characteristics listed in Appendix Table A.2.
54
In addition, 12.8 percent of unbanked households selected other method in 2017, which is substantially higher than the percentage of underbanked (3.9 percent) and
fully banked (2.0 percent) households that selected other method. (See Appendix Table E.4.)
55
47
2017 FDIC National Survey of Unbanked and Underbanked
Households
8. Credit
Building on the 2015 survey, which introduced questions
about small-dollar bank credit, the 2017 survey included new
questions to capture the full range of credit products that are
likely reported to the major credit bureaus (i.e., mainstream
credit). Specifically, the 2015 survey asked households
whether, in the past 12 months, they had a credit card from
Visa, MasterCard, American Express, or Discover (i.e., credit
card) or a personal loan or line of credit from a bank (i.e., bank
personal loan). Additional questions in the 2017 survey asked
households whether, in the past 12 months, they had a store
credit card; an auto loan; a student loan; a mortgage, home
equity loan, or home equity line of credit (HELOC); or other
personal loans or lines of credit from a company other than a
bank (i.e., other mainstream nonbank).56
months. Households that did not have mainstream credit in
the past 12 months likely did not have a credit score, which
could make it more difficult to obtain mainstream credit
should a credit need arise.
The 2017 survey also retained questions from the 2015 survey
that provide information on credit applications and potential
indicators of creditworthiness (i.e., credit characteristics).
Specifically, households were asked whether, in the past 12
months, they applied for a credit card or bank personal loan
(i.e., applied); were denied a credit card or bank personal
loan, or not given as much credit as they applied for (i.e.,
denied); thought about applying for a credit card or bank
personal loan but did not because they thought they might
be turned down (i.e., felt discouraged about applying); or fell
behind on bills.
A household is considered to have used mainstream credit if it used any of the above credit products in the past 12
Figure 8.1 Use of Mainstream Credit Products, 2017 (Percent)
Credit card
68.7
Store credit card
41.6
Mortgage, home equity loan, or HELOC
33.8
Auto loan
32.3
Student loan
16.6
Bank personal loan
Other mainstream nonbank
No mainstream credit
6.9
2.1
19.7
Other mainstream nonbank credit includes finance company loans and purchase loans or lines of credit from retailers. This category does not include credit AFS.
56
48 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Mainstream Credit Product Use
Figure 8.1 presents the shares of households in 2017 that
used each mainstream credit product.57 Credit cards were the
most common (68.7 percent of households had a credit card
from Visa, MasterCard, American Express, or Discover, and
41.6 percent had a store credit card), followed by mortgages, home equity loans, or HELOCs; and auto loans. Student
loans, bank personal loans, and other mainstream nonbank
credit were much less common.
Mainstream Credit Product Use by Banking Status
and Household Characteristics
Table 8.1 shows the shares of households in 2017 that used
each mainstream credit product by banking status and household characteristics. Use of each mainstream credit product
was much lower among unbanked households, relative to
underbanked and fully banked households. For example,
only 7.2 percent of unbanked households had a credit card,
compared with 60.0 percent of underbanked households and
76.3 percent of fully banked households. Use of mainstream
credit products also varied widely across socioeconomic and
demographic groups. In general, lower-income households,
less-educated households, the youngest and oldest households, black and Hispanic households, and working-age
disabled households were less likely to use most mainstream
credit products.
Credit usage patterns varied somewhat by product type. For
example, while similar shares of black and white households
had student loans, use of the remaining mainstream credit
products was generally lower among black households.
Credit Characteristics by Banking Status and
Household Characteristics
Table 8.2 presents the proportions of households in 2017 that
applied for a credit card or bank personal loan by banking
status and selected household characteristics. Overall, 14.1
percent of households applied for a credit card or bank
personal loan in the past 12 months. Unbanked households applied at a substantially lower rate (3.0 percent) than
underbanked (18.0 percent) and fully banked households
(14.0 percent). Certain segments of the population, including
lower-income households, less-educated households, older
households, and black and Hispanic households, also applied
at lower rates than other segments.
Table 8.2 also shows the shares of households in 2017 that
applied for a credit card or bank personal loan and were denied, that felt discouraged about applying, or that fell behind
on bills. Overall, 19.5 percent of households were denied a
credit card or bank personal loan (conditional on having applied), 5.6 percent felt discouraged about applying, and 14.5
percent fell behind on bills. Lower-income households, working-age disabled households, and households with volatile
income were more likely to have been denied (conditional on
having applied), to have felt discouraged about applying, or to
have fallen behind on bills.
Changes in Credit Card Ownership and Credit
Characteristics
Table 8.3 shows that credit card ownership and a few of the
credit characteristics changed somewhat from 2015 to 2017.
Credit card ownership increased slightly from 66.5 percent in
2015 to 68.7 percent in 2017.58 The share of households that
felt discouraged about applying for a credit card or bank personal loan declined from 2015 to 2017. This decline can be
attributed primarily to changes in income and other characteristics of U.S. households between 2015 and 2017.59 Finally,
the share of households that fell behind on bills decreased.
Share of Households With No Mainstream Credit by
Banking Status and Household Characteristics
A positive credit history can promote financial resiliency and
overall financial health. Households with an insufficient credit
history, however, likely face substantially reduced access to
mainstream credit. A positive credit history also facilitates
large purchases that may not be feasible without credit, such
as a house or car. Finally, lack of a credit history may affect
employment and rental housing opportunities, as prospective
employers and landlords often rely on credit reports as part of
the application process.
As shown in Table 8.1, one in five (19.7 percent) households
in 2017 had no mainstream credit in the past 12 months and
likely did not have a credit score. Households without a credit
score may be “credit invisible,” meaning that no one in the
household has a record at one of the credit bureaus. Alternatively, a household member may have a record at one of
the credit bureaus but not have sufficient credit history to be
scored. At least one active trade line in the past six months is
generally required to generate a credit score.
The analysis presented in this section excludes 3,622 observations (representing roughly 14.2 million households) with missing information on whether the household
used one or more mainstream credit products; used credit AFS; applied for, was denied, or felt discouraged about applying for a credit card or bank personal loan; or
fell behind on bills. In Table 8.3, the 2015 estimates are identical to those in the 2015 report.
57
The share of households with a bank personal loan was 9.8 percent in 2015 and 6.9 percent in 2017; however, these estimates are not necessarily comparable across
time because of changes in question wording and placement between the 2015 and 2017 surveys. It is possible that some respondents included more products as
personal loans or lines of credit from a bank, such as home equity loans or HELOCs, in the 2015 survey than in the 2017 survey. See Appendix 2 for additional details.
58
After accounting for changes in the household characteristics listed in Appendix Table A.2, the remaining decline from 2015 to 2017 in the share of households that
felt discouraged about applying was not statistically significant.
59
49
Table 8.1 Use of Mainstream Credit Products by Banking Status and Selected Household Characteristics, 2017
For all households, row percent
Credit card
(Percent)
Store
credit card
(Percent)
Mortgage,
home
equity
loan, or
HELOC
(Percent)
Auto loan
(Percent)
Student
loan
(Percent)
Bank
personal
loan
(Percent)
Other
mainstream
nonbank
(Percent)
No
mainstream
credit
(Percent)
68.7
41.6
33.8
32.3
16.6
6.9
2.1
19.7
Unbanked
7.2
4.0
3.4
5.7
4.5
1.2
1.0
80.2
Underbanked
60.0
37.9
26.3
35.5
20.6
8.4
4.3
21.9
Fully banked
76.3
45.7
38.5
33.7
16.5
7.0
1.6
14.1
Characteristics
All
Banking status
Family income
Less than $15,000
31.3
16.7
6.7
9.4
8.4
2.1
1.4
56.2
$15,000 to $30,000
48.6
25.8
13.9
16.0
9.3
3.4
1.8
35.8
$30,000 to $50,000
64.1
37.2
23.7
28.0
14.0
5.6
2.3
20.1
$50,000 to $75,000
77.0
47.4
37.0
37.7
17.7
7.6
2.5
11.2
At least $75,000
88.8
56.7
56.2
47.0
23.6
10.5
2.1
4.3
No high school diploma
33.3
19.3
12.8
14.2
2.9
2.6
1.7
53.0
High school diploma
57.7
36.0
25.8
27.2
8.4
5.9
1.9
28.1
Some college
68.4
42.9
33.5
34.8
18.1
7.6
2.5
17.2
College degree
86.0
50.3
45.3
38.6
24.7
8.2
1.9
7.2
15 to 24 years
56.1
26.2
12.8
31.4
31.4
4.5
2.2
24.6
25 to 34 years
67.9
35.4
29.7
40.4
32.8
7.4
2.1
18.5
35 to 44 years
69.1
41.3
43.0
42.0
24.5
8.6
2.5
18.0
45 to 54 years
69.3
45.7
45.1
39.6
16.4
8.2
2.4
17.9
55 to 64 years
69.6
45.7
39.4
29.8
10.0
7.6
2.2
19.2
65 years or more
70.6
43.0
22.3
17.2
2.8
4.5
1.4
22.4
Black
47.9
28.2
19.3
23.7
17.8
5.7
2.8
36.0
Hispanic
53.9
35.3
24.6
29.0
14.3
5.2
1.4
31.5
Asian
80.8
42.5
36.1
32.5
14.0
5.6
0.9
13.8
White
75.3
45.7
38.5
34.7
17.0
7.6
2.1
14.4
Other
52.2
31.2
28.1
32.7
15.6
8.5
2.9
28.3
Disabled, age 25 to 64
43.1
27.6
22.9
21.1
10.5
6.0
3.2
40.4
Not disabled, age 25 to 64
72.7
44.3
41.8
40.1
21.8
8.2
2.2
15.3
Income was about the
same each month
69.2
41.5
33.5
31.4
15.7
6.1
1.8
19.9
Income varied somewhat
from month to month
68.2
43.6
36.8
36.7
20.1
9.7
2.8
18.1
Income varied a lot from
month to month
62.8
35.7
28.1
31.0
19.2
9.8
4.2
23.5
U.S.-born
69.8
42.5
34.9
33.3
17.5
7.3
2.2
18.5
Foreign-born citizen
70.2
42.7
35.3
28.8
14.6
5.4
1.2
19.9
Foreign-born noncitizen
53.3
28.9
18.6
23.5
7.1
3.2
1.5
35.9
Education
Age group
Race/Ethnicity
Disability status
Monthly income volatility
Nativity
Note: See Appendix Table F.1 for estimates by other household characteristics.
50 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Table 8.2 Credit Characteristics by Banking Status and Selected Household Characteristics, 2017
For all households, row percent
Applied
(Percent)
Denied
(Percent)
Felt discouraged
about applying
(Percent)
Fell behind
on bills
(Percent)
Denied, conditional
on applying
(Percent)
14.1
2.8
5.6
14.5
19.5
Unbanked
3.0
1.7
8.5
39.9
NA
Underbanked
18.0
6.2
13.2
29.1
34.2
Fully banked
14.0
1.9
3.3
8.3
13.7
Less than $15,000
6.5
2.4
8.5
27.3
36.7
$15,000 to $30,000
8.4
3.1
7.3
22.9
36.7
$30,000 to $50,000
11.6
3.1
7.0
16.7
26.3
$50,000 to $75,000
15.2
2.9
5.4
12.1
18.8
At least $75,000
20.2
2.5
3.1
6.3
12.4
No high school diploma
5.4
1.5
6.2
22.3
28.3
High school diploma
10.2
2.1
5.7
17.3
20.5
Some college
14.5
3.8
7.3
18.1
26.2
College degree
18.9
2.7
4.0
7.6
14.2
15 to 24 years
19.0
5.5
9.8
20.5
28.8
25 to 34 years
19.9
4.1
9.2
18.0
20.8
35 to 44 years
16.9
3.5
7.3
19.0
20.5
45 to 54 years
16.1
3.5
6.2
17.2
22.0
55 to 64 years
12.4
2.1
4.1
13.6
17.2
65 years or more
7.4
0.7
1.9
6.5
9.3
Black
10.2
3.9
10.1
27.2
37.6
Hispanic
12.3
3.2
7.0
18.9
25.7
Asian
20.7
2.7
3.9
7.2
12.9
White
14.8
2.4
4.5
11.3
16.3
Other
12.6
4.4
7.5
24.1
NA
Disabled, age 25 to 64
11.3
3.7
9.0
30.9
32.8
Not disabled, age 25 to 64
16.9
3.2
6.3
14.8
19.1
Income was about the same each month
13.1
2.3
4.6
11.9
17.7
Income varied somewhat from month to month
18.0
4.1
8.5
21.0
22.6
Income varied a lot from month to month
17.9
5.5
11.4
35.5
30.6
U.S.-born
14.0
2.6
5.6
14.5
18.9
Foreign-born citizen
15.1
2.8
4.9
12.7
18.4
Foreign-born noncitizen
15.0
4.2
6.1
16.5
27.7
Characteristics
All
Banking status
Family income
Education
Age group
Race/Ethnicity
Disability status
Monthly income volatility
Nativity
Notes: NA indicates that the sample size was too small to produce a precise estimate. See Appendix Tables F.5 – F.9 for estimates by other household
characteristics.
51
Table 8.3 Credit Card Ownership and Credit Characteristics by Year
For all households
2015
(Percent)
2017
(Percent)
Difference
(2017-2015)
Credit card
66.5
68.7
2.2*
Applied
13.9
14.1
0.2
Denied
2.8
2.8
0.0
Felt discouraged about applying
6.1
5.6
-0.5*
Fell behind on bills
16.9
14.5
-2.4*
Denied, conditional on applying
20.0
19.5
-0.5
Notes: Asterisks indicate differences that are statistically significant at the 10 percent level. See Appendix Tables F.4 – F.9 for estimates by banking status and
household characteristics and for selected confidence intervals.
Figure 8.2 No Mainstream Credit by Household Race and Ethnicity and Income Level, 2017 (Percent)
Less than $15,000
48.2
$15,000 to $30,000
48.7
47.9
29.1
$30,000 to $50,000
16.2
$50,000 to $75,000
At least $75,000
63.4
8.5
3.5
6.8
27.9
28.5
18.9
18.5
9.5
Black
Hispanic
White
Note: To simplify the figure, estimates for Asian households and for households of other races and ethnicities are not shown.
52 | 2017 FDIC National Survey of Unbanked and Underbanked Households
67.8
Differences in the share of households with no mainstream
credit by banking status were striking. Four in five (80.2
percent) unbanked households had no mainstream credit,
compared with 21.9 percent of underbanked households
and 14.1 percent of fully banked households. The share
of households with no mainstream credit also varied
substantially across socioeconomic and demographic groups.
Lower-income households, less-educated households, black
and Hispanic households, working-age disabled households,
and foreign-born, noncitizen households were more likely
not to have mainstream credit. Differences by income and
education were especially pronounced. For example, 56.2
percent of households with less than $15,000 in income
had no mainstream credit, compared with only 4.3 percent
of households with income of $75,000 or more. Similarly,
53.0 percent of households with no high school diploma had
no mainstream credit, compared with only 7.2 percent of
households with a college degree.
likely not to have mainstream credit. Racial and ethnic differences in bank account ownership and socioeconomic and
demographic characteristics beyond income can account for
some, but not all, of the racial and ethnic differences in the
likelihood of not having mainstream credit.
The youngest and oldest age groups were more likely not
to have mainstream credit, although after accounting for
differences in income across age groups, households aged 15
to 24 years were less likely than the other age groups not to
have mainstream credit.
Share of Households With No Mainstream Credit
by Geography
The share of households with no mainstream credit varied
across regions of the United States. Almost one in four (23.8
percent) households in the South had no mainstream credit,
compared with 18.2 percent in the Northeast, 17.3 percent in
the West, and 16.5 percent in the Midwest. Figure 8.3 shows
that the share of households with no mainstream credit varied
widely across states, ranging from 8.1 percent in Minnesota
to 37.7 percent in Mississippi. (See Appendix Tables F.2 and
F.3 for detailed state- and MSA-level estimates.)
Differences by race and ethnicity were also substantial: 36.0
percent of black households and 31.5 percent of Hispanic
households had no mainstream credit, compared with 14.4
percent of white households. As shown in Figure 8.2, at all
income levels, black and Hispanic households were more
Figure 8.3 No Mainstream Credit by State, 2017 (Percent)
WA
ND
ND
MT
OR
MN
SD
ID
MI
WY
UT
IL
CO
KS
CA
AZ
OK
NM
MO
WV
KY
TN
LA
AL
MD
VA
MA
CT
NJ
RI
DE
DC
NC
AR
MS
TX
PA
OH
IN
VT
NH
NY
IA
NE
NV
ME
WI
SC
GA
Less than 14.0
14.0 to 16.4
FL
AK
HI
16.4 to 20.0
20.0 to 24.0
At least 24.0
53
Figure 8.4 Interest in Having Credit Among Households With No Mainstream Credit, 2017 (Percent)
Showed interest in having credit
Applied
15.8
2.5
Felt discouraged about applying
6.4
Used credit AFS
10.1
Stayed current on bills
76.3
Figure 8.5 Interest in Having Credit Among Households With No Mainstream Credit by Household Race and Ethnicity, 2017
(Percent)
19.6
16.2
14.1
Showed interest in having credit
Applied
2.2
3.3
2.4
8.8
8.2
Felt discouraged about applying
4.7
12.5
Used credit AFS
8.4
9.8
68.5
Stayed current on bills
76.2
79.7
Black
Hispanic
Note: To simplify the figure, estimates for Asian households and for households of other races and ethnicities are not shown.
54 | 2017 FDIC National Survey of Unbanked and Underbanked Households
White
Interest in Having Credit Among Households With
No Mainstream Credit
Two reasons why households may not have mainstream
credit are that they are not interested in having credit or that
they do not appear creditworthy. For the purposes of this
report, we consider a household to have shown interest in
having credit if the household applied for a credit card or
bank personal loan, felt discouraged about applying, or used
credit AFS.60
As indicated in Figure 8.4, approximately one in six (15.8
percent) households with no mainstream credit in 2017
showed interest in having credit. Among households with
no mainstream credit, 2.5 percent applied for a credit card
or bank personal loan, 6.4 percent felt discouraged about
applying, and 10.1 percent used credit AFS.61
Staying current on bills is one potential indicator of
creditworthiness. As shown in Figure 8.4, about three in four
(76.3 percent) households with no mainstream credit stayed
current on bills in the past 12 months. Among households
with no mainstream credit that showed interest in having
credit, roughly half (46.7 percent) stayed current on bills.
While staying current on bills is an imperfect measure of
creditworthiness, it nevertheless provides some insight into
these households’ financial situation.
Figure 8.5 shows interest in having credit and the likelihood
of staying current on bills among households with no
mainstream credit in 2017 by race and ethnicity. Among
these households, 19.6 percent of black households showed
interest in having credit, compared with 16.2 percent of
Hispanic households and 14.1 percent of white households.62
Among households with no mainstream credit, the share that
stayed current on bills was similar for Hispanic and white
households, while the share was lower for black households.
Unmet Demand for Mainstream Small-Dollar Credit
Households may use certain credit products, including
credit cards, bank personal loans, and credit AFS, to meet
their small-dollar credit needs. Some households may have
small-dollar credit needs that are not fully met by mainstream
financial institutions. As in the 2015 report, we classify a
household as having unmet demand for mainstream smalldollar credit if the household applied for and was denied a
credit card or bank personal loan, felt discouraged about
applying, or used credit AFS. Applying this convention, 12.9
percent of households had unmet demand for mainstream
small-dollar credit in 2017, compared with 13.7 percent in
2015. The decline in the share of households with unmet
demand from 2015 to 2017 is consistent with the declines
in the shares of households that used credit AFS or that felt
discouraged about applying for a credit card or bank personal
loan. Among households with unmet demand, 57.2 percent
stayed current on bills in 2017, up slightly from 52.5 percent
in 2015.
This definition is an approximation and likely does not capture all households that have shown interest in having credit. For example, households may have applied
for or have felt discouraged about applying for other credit products, such as auto loans or student loans.
60
For comparison, 24.8 percent of households with mainstream credit applied for a credit card or bank personal loan, felt discouraged about applying, or used credit
AFS: 17.0 percent applied for a credit card or bank personal loan, 5.4 percent felt discouraged about applying, and 6.8 percent used credit AFS.
61
The difference between Hispanic and white households in the share that showed interest in having credit was not statistically significant. Further, after accounting
for differences between black and white households in the household characteristics listed in Appendix Table A.2, the remaining difference between black and white
households in the share that showed interest in having credit was not statistically significant.
62
55
2017 FDIC National Survey of Unbanked and Underbanked
Households
9. How Households Conduct Their Financial Transactions in a Typical Month
As in the 2015 survey, the 2017 survey included a number
of questions about the ways households pay bills (for things
like mortgage, rent, utilities, or child care) and receive income
(from work, retirement, government benefits, or other sources)
in a typical month. The goal of these questions was to learn
more about the extent to which households use bank and
other methods to meet their financial transactions needs in a
typical month.63
For the purposes of this report, the following methods of
paying bills are classified as bank methods: electronic payment from a bank account, personal check drawn on a bank,
debit card linked to a bank account, credit card, and cashier’s
check or money order purchased at a bank. Other bill payment methods include nonbank money orders, prepaid cards,
and cash.
Similarly, the following methods of receiving income are classified as bank methods: direct deposit into a bank account,
and paper check or money order if the household had a
bank account and did not go to a nonbank check casher in
a typical month. Other methods of receiving income include
cash, direct deposit onto a prepaid card, and paper check or
money order (for households that were unbanked or that used
a nonbank check casher to get the funds).64
National Estimates
The great majority of U.S. households used bank methods to
pay bills in a typical month, consistent with the fact that most
U.S. households have a bank account.65 As illustrated in Figure 9.1, the most widely used method in 2017 was electronic
payment from a bank account (68.4 percent). Use of personal
checks was also common (61.3 percent), while use of other
bill payment methods, such as cash, nonbank money orders,
and prepaid cards, was substantially lower. Overall, 93.8 percent of households used at least one bank method to pay bills
in a typical month, and 78.2 percent used only bank methods
(i.e., they used a bank method and did not use cash, nonbank
money orders, prepaid cards, or other).
Figure 9.2 shows methods used to receive income in a typical
month. The most prevalent method was by far direct deposit
into a bank account (86.7 percent). Approximately one in four
households (27.6 percent) received income by paper check
or money order. Of the households that used this method,
7.0 percent (or 1.9 percent of all households) used a nonbank
check casher to get the funds in a typical month.66 Less commonly used ways of receiving income were cash and direct
deposit onto a prepaid card. Overall, 93.2 percent of households used at least one bank method to receive income in a
typical month, and 84.1 percent used only bank methods.
For the 2017 survey, a few changes were made to the questions on bill payment and income receipt methods. Retirement was explicitly added as a source of
income in the introduction to the questions on income receipt. To accommodate new questions in the 2017 survey, a question that asked households to choose the
primary (i.e., most common) method of bill payment in a typical month was dropped, and the remaining questions on bill payment and income receipt methods were
streamlined. Finally, in the 2015 and 2017 surveys households could volunteer that they did not pay bills, but in the 2017 survey households that did not select a bill
payment method and that did not volunteer that they did not pay bills were explicitly asked whether they paid bills (a similar question was asked for income receipt).
See Appendix 2 for additional details.
63
The distinction between bank and other methods is not always straightforward. The approach used in this report is to classify a method as a bank method if a
bank is likely to be directly involved in the transaction, at least from the household’s perspective. Use of prepaid cards to pay bills or receive income is treated as
other because, in most cases, consumers do not obtain the card directly from a bank. Similarly, use of cash to pay bills or receive income is considered to be other,
although, in some cases, the cash may have been obtained directly from a bank account (particularly among banked households). Unbanked households that received
income through paper check or money order and did not typically use a nonbank check casher are also classified as other, although it is possible that in at least some
of these cases, the households may have gone to a bank to receive the funds.
64
The analysis presented in this section excludes 4,514 observations (representing 17.8 million households) with missing information on use of prepaid cards or
nonbank money orders or check cashers in the past 12 months, on methods used to pay bills or receive income, or where the household indicated that it did not pay
bills or receive income in a typical month.
65
This does not necessarily mean that only 1.9 percent of households used a nonbank check casher in a typical month. Households may use nonbank check cashers
to handle paper checks or money orders that they do not think of as income.
66
56 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Figure 9.1 Methods Used to Pay Bills in a Typical Month, 2017 (Percent)
Electronic payment from bank
68.4
Personal check
61.3
Debit card
47.3
Credit card
24.8
Bank money order
5.7
Cash
15.9
Nonbank money order
Prepaid card
6.9
2.3
Other
1.2
Did not select a method
0.5
Any bank method
93.8
Only bank methods
78.2
Figure 9.2 Methods Used to Receive Income in a Typical Month, 2017 (Percent)
Direct deposit into bank account
86.7
Paper check or money order
27.6
Cash
Direct deposit onto prepaid card
7.9
3.4
Other
1.8
Nonbank check casher
1.9
Did not select a method
1.6
Any bank method
Only bank methods
93.2
84.1
57
While the 2017 estimates in Figures 9.1 and 9.2 are qualitatively similar to the estimates presented in the 2015 report,
changes to the administration of the survey instrument make
direct comparisons difficult.67 As shown in Appendix Table
G.1, the rank ordering of bill payment methods by prevalence
was unchanged from 2015 to 2017, and the same was true of
income receipt methods.68 However, use of paper instruments
to pay bills and receive income declined somewhat between
2015 and 2017, while use of electronic methods increased.
Although personal checks remained the second-most prevalent method of paying bills, the proportion of households
that used this method decreased from 2015 to 2017. Over the
same period, the proportions that used electronic payments
from a bank account, debit cards, or credit cards increased.69
Likewise, the proportion of households that received income
by paper check or money order decreased from 2015 to 2017,
while the proportion that received income through direct deposit into a bank account increased.70
Bill Payment and Income Receipt Methods by
Household Characteristics
As in 2015, most U.S. households used banks to handle their
financial transactions in a typical month, although certain
segments of the population were less likely to do so.71 Panel
A of Table 9.1 illustrates differences by income in the methods
households used to pay bills in 2017. Lower-income households were substantially less likely to use bank methods and
more likely to use other methods, such as cash and nonbank
money orders. Most notably, use of electronic payments from
a bank account varied sharply by income, ranging from 37.8
percent of households with less than $15,000 in income to
85.6 percent of households with income of $75,000 or more.
Moreover, 33.7 percent of households with less than $15,000
in income paid bills in cash, compared with only 7.0 percent
of households with income of $75,000 or more.
Lower-income households were also substantially less likely
to receive income using bank methods. As shown in panel B
of Table 9.1, while direct deposit into a bank account was the
most prevalent method of receiving income at each income
level, its use was more common among higher-income households. Close to 70 percent (69.2 percent) of households with
less than $15,000 in income received income through direct
deposit into a bank account, compared with 94.9 percent of
households with income of $75,000 or more. The proportion
of households that received income by paper check or money
order was fairly similar across income groups. However,
lower-income households were more likely to use a nonbank
check casher to get those funds.
Examining differences across other household characteristics
revealed that overall use of bank methods to pay bills and
receive income in a typical month was less prevalent among
less-educated households, younger households, and black
and Hispanic households. Even within these groups, the proportion of households that used bank methods was still high
relative to the proportions that used other methods.72
Specifically, in the 2015 and 2017 surveys households could volunteer that they did not pay bills, but in the 2017 survey households that did not select a bill
payment method and that did not volunteer that they did not pay bills were explicitly asked whether they paid bills (a similar question was asked for income receipt).
The analysis in the 2015 report excludes households that volunteered that they did not pay bills or receive income, while the analysis in this report also excludes
households that indicated that they did not pay bills or receive income when explicitly asked. Moreover, as discussed in Appendix 1 of the 2015 report, because of
an issue with the administration of the 2015 survey instrument, information on use of prepaid cards to receive income is missing for many unbanked households. The
2015 estimates of income received through direct deposit onto a prepaid card or using other methods incorporate imputed values for these households.
67
To more directly compare the 2015 and 2017 estimates, in Appendix Table G.1 households that indicated that they did not pay bills or receive income when explicitly
asked are not excluded from the 2017 estimates, as they are elsewhere in this report.
68
Based on the 2017 estimates shown in Appendix Table G.1, use of personal checks to pay bills decreased from 61.2 percent in 2015 to 59.8 percent in 2017. In
contrast, use of electronic payments from a bank account increased from 64.3 percent to 66.5 percent, use of debit cards increased from 39.7 percent to 46.2 percent,
and use of credit cards increased from 21.3 percent to 24.3 percent.
69
Based on the 2017 estimates shown in Appendix Table G.1, the proportion of households that received income by paper check or money order decreased from 29.1
percent in 2015 to 26.7 percent in 2017, while the proportion that received income through direct deposit into a bank account increased from 81.3 percent to 84.0
percent.
70
Differences across households in the methods used to pay bills and receive income may be attributable to a number of factors, some of which may be outside of the
household’s control, such as the ways employers disburse earnings (e.g., availability of direct deposit or use of payroll cards) or in the types of payment instruments
required by payees.
71
See Appendix Tables G.4 – G.11 for estimated use of bill payment and income receipt methods in a typical month by selected household characteristics. As
with household income, there are differences by other household characteristics in the specific methods used to pay bills and receive income. For example, older
households were less likely than younger households to pay bills electronically from a bank account and more likely to pay bills by personal check.
72
58 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Table 9.1 Methods Used to Pay Bills and Receive Income in a Typical Month by Household Income Level, 2017
For all households that paid bills and received income in a typical month, column percent
All
Less than
$15,000
$15,000 to
$30,000
$30,000 to
$50,000
$50,000 to
$75,000
At least
$75,000
A. Paying bills (Percent)
Electronic payment from bank
68.4
37.8
48.6
62.4
75.2
85.6
Personal check
61.3
44.1
56.1
62.2
64.4
66.6
Debit card
47.3
36.7
44.6
50.5
53.1
46.9
Credit card
24.8
13.4
16.0
21.0
25.6
33.7
Bank money order
5.7
10.0
8.4
6.5
4.8
3.3
Cash
15.9
33.7
26.3
19.1
11.1
7.0
Nonbank money order
6.9
17.4
13.0
8.3
4.3
1.6
Prepaid card
2.3
6.1
4.3
2.6
1.3
0.5
Other
1.2
2.9
2.2
1.0
0.8
0.7
Did not select a method
0.5
1.2
0.7
0.5
0.5
0.2
Any bank method
93.8
75.2
87.8
94.4
98.2
99.3
Only bank methods
78.2
52.2
64.0
74.6
84.1
90.8
Direct deposit into bank account
86.7
69.2
76.0
85.5
91.0
94.9
Paper check or money order
27.6
23.8
28.7
28.7
27.3
27.9
Cash
7.9
12.4
10.3
8.8
6.7
5.7
B. Receiving income (Percent)
Direct deposit onto prepaid card
3.4
8.7
5.0
3.3
2.1
1.8
Other
1.8
4.4
2.3
1.7
1.2
1.3
Nonbank check casher
1.9
4.4
4.5
2.3
0.9
0.5
Did not select a method
1.6
4.6
2.4
1.2
1.3
0.8
Any bank method
93.2
75.8
86.0
93.9
97.4
98.8
Only bank methods
84.1
66.4
76.5
83.6
88.4
90.5
Bill Payment and Income Receipt Methods by
Banking Status
As in 2015, unbanked households in 2017 paid bills and received income primarily using methods outside of the banking
system. As shown in Table 9.2, approximately two-thirds
of unbanked households paid bills in 2017 using cash, the
most prevalent method. Nonbank money orders and prepaid
cards were the next two most prevalent methods of paying
bills. Unbanked households also received income in a variety
of ways, but the most prevalent method was paper check
or money order, followed by cash and direct deposit onto a
prepaid card.73
Underbanked households, on the other hand, used banks
extensively to handle their financial transactions. In fact, 94.0
percent of underbanked households used at least one bank
method to pay bills, a share that is almost as high as the
estimate for fully banked households (98.4 percent). Electronic payment from a bank account was the most widely
used method of paying bills among both underbanked and
fully banked households. Relative to the fully banked, use of
personal checks was lower among underbanked households
and use of debit cards was higher.
The key difference between underbanked and fully banked
households is that, in addition to using bank methods, the
underbanked also widely used other methods to pay bills.
Use of cash or nonbank money orders was substantially
higher among underbanked households, compared with the
fully banked.74 As a result, the proportion of households that
used only bank methods to pay bills was much lower among
the underbanked.
About half of the unbanked households that received income by paper check or money order (or 23.7 percent of all unbanked households) used a nonbank check
casher to get the funds in a typical month. The remaining unbanked households that received income by paper check or money order did not use a nonbank check
casher to get those funds. We do not directly observe how these households obtained the funds from the income received by paper check or money order.
73
74
By definition, fully banked households did not use nonbank money orders (or any other AFS asked about in the survey) in the past 12 months.
59
Underbanked households were also almost as likely as fully
banked households to use bank methods to receive income
in a typical month. Direct deposit into a bank account was by
far the most prevalent method of receiving income, both for
underbanked and fully banked households.
Patterns were similar when looking at bill payment together
with income receipt. As shown in Table 9.3, unbanked households primarily operated outside of the banking system, using
cash, transaction AFS, prepaid cards, or other methods. In
contrast, almost all (99.0 percent) underbanked households
used at least one bank method to pay bills or receive income
in a typical month, and nearly half (48.5 percent) used only
bank methods to pay bills and receive income. These findings
confirm that unbanked households did not participate in the
mainstream financial system to the same extent as underbanked households, at least when handling these financial
transactions.
Table 9.2 Methods Used to Pay Bills and Receive Income in a Typical Month by Banking Status, 2017
For all households that paid bills and received income in a typical month, column percent
All
Unbanked
Underbanked
Fully banked
Electronic payment from bank
68.4
2.5
67.2
73.0
Personal check
61.3
1.2
52.0
67.8
Debit card
47.3
3.1
63.1
45.9
Credit card
24.8
8.4
25.0
25.8
Bank money order
5.7
13.0
11.8
3.5
Cash
15.9
66.1
26.2
9.8
Nonbank money order
6.9
39.1
24.2
0.0
Prepaid card
2.3
22.1
4.0
0.5
Other
1.2
8.0
1.3
0.7
A. Paying bills (Percent)
Did not select a method
0.5
3.3
0.3
0.4
Any bank method
93.8
22.7
94.0
98.4
Only bank methods
78.2
6.2
56.6
88.8
Direct deposit into bank account
86.7
5.6
86.6
92.0
Paper check or money order
27.6
45.4
30.8
25.6
Cash
7.9
26.5
10.5
6.0
Direct deposit onto prepaid card
3.4
23.3
5.0
1.6
Other
1.8
10.6
1.9
1.2
Nonbank check casher
1.9
23.7
3.8
0.0
Did not select a method
1.6
10.5
1.3
1.1
Any bank method
93.2
5.6
95.3
98.2
Only bank methods
84.1
2.6
80.0
90.5
B. Receiving income (Percent)
Table 9.3 Joint Methods of Paying Bills and Receiving Income in a Typical Month by Banking Status, 2017
For all households that paid bills and received income in a typical month, column percent
All
Unbanked
Underbanked
Fully banked
Any bank method
96.0
24.7
99.0
99.8
Cash
21.0
70.6
32.6
14.6
Prepaid card
4.2
29.7
7.0
1.8
Transaction AFS
7.8
47.9
26.7
0.0
Other or none selected
4.8
27.3
4.6
3.4
Only bank methods
70.6
1.1
48.5
81.3
Notes: “Any bank method” includes households that used at least one bank method to pay bills or receive income. “Transaction AFS” includes households that used
a nonbank money order to pay bills or that used a nonbank check casher to get the money from income received by paper check or money order. “Other or none
selected” includes households that indicated they used other methods for bill payment or income receipt, or that did not select a method of bill payment or income
receipt. “Only bank methods” includes households that used bank methods to pay bills and receive income, and did not use any other methods.
60 | 2017 FDIC National Survey of Unbanked and Underbanked Households
2017 FDIC National Survey of Unbanked and Underbanked
Households
10. Measuring Economic Inclusion
A primary goal of the FDIC National Survey of Unbanked and
Underbanked Households is to assess the inclusiveness of
the U.S. banking system. Specifically, the survey is used to
estimate the proportion of households that do not have an
account at a federally insured depository institution (i.e., the
unbanked rate) and the proportion that have an account but
go outside of the banking system to meet their financial needs
(i.e., the underbanked rate). As consumer financial product
markets evolve and new products mature, measurement of
the unbanked and underbanked may require updating to
reflect such changes and to better assess the inclusiveness of
the banking system. For these reasons, this section explores
measurement of the unbanked and underbanked and considers refinement of the unbanked and underbanked definitions.
Measurement of the Unbanked
In this report and since the survey was first conducted in
2009, a household is categorized as unbanked if no one in the
household has a checking or savings account. As discussed
in section 5, some consumers use general purpose reloadable
prepaid cards to address their financial transactions needs.
These cards may offer many of the same features as checking
accounts, including a safe place to receive and store funds,
the ability to withdraw cash from ATMs, and bill payment services. Further, a household that obtains a prepaid card from a
bank may benefit from having a relationship with a bank, such
as expanded access to other banking products and services.
As a result, unbanked households that use prepaid cards obtained from banks could be considered banked. If they were,
the unbanked rate in 2017 would fall slightly from 6.5 percent
to 6.4 percent.
loans, rent-to-own services, pawn shop loans, or auto title
loans.75 This underbanked definition does not incorporate
intensity of AFS use: some underbanked households may
routinely use AFS, while others may do so only sporadically.
It also considers a wide range of AFS, including transaction
and credit products and services. The costs and availability of
these products and services vary, both in absolute terms and
relative to comparable services offered by banks. As a result,
households categorized as underbanked in this report are a
fairly broad group, with a variety of experiences and levels of
engagement with the banking system.
To better understand the financial behaviors of the underbanked and their engagement with the banking system, the
remainder of this section segments underbanked households
into two groups based on whether they used only bank methods to pay bills and receive income in a typical month.76 We
then explore the socioeconomic and demographic characteristics of households in each group, as well as how households in each group access their bank accounts and meet
their transactions, savings, and credit needs. For context, we
also compare the groups to the unbanked and fully banked.
In 2017, approximately half (48.6 percent) of underbanked
households used only bank methods to pay bills and receive
income in a typical month, which we denote as underbanked
group 1.77 The remaining 51.4 percent of underbanked
households did not exclusively use bank methods to pay bills
and receive income in a typical month, which we denote as
underbanked group 2.
Household Characteristics
Measurement of the Underbanked
In this report and since 2013, a household is classified as
underbanked if it has a checking or savings account and
used one of the following products or services from an AFS
provider in the past 12 months: money orders, check cashing,
international remittances, payday loans, refund anticipation
As shown in Table 10.1, the socioeconomic and demographic
characteristics of households in underbanked group 1 were
generally similar to the characteristics of the fully banked.
Relative to households in underbanked group 1 and to the
fully banked, households in underbanked group 2 had lower
income and educational attainment; were more likely to be
Underbanked definitions were different in the 2009 and 2011 reports, in part because of differences in the sets of AFS asked about in the 2009 and 2011 surveys.
International remittances were first asked about in the 2011 survey, and auto title loans were first asked about in the 2013 survey.
75
The analysis presented in the remainder of this section excludes 5,402 observations (representing roughly 21.1 million households) with missing information on
whether the household saved for unexpected expenses or emergencies; used one or more mainstream credit products; used credit AFS; applied for, was denied, or
felt discouraged about applying for a credit card or bank personal loan; or fell behind on bills. The analysis also excludes households with missing information on use
of prepaid cards or nonbank money orders or check cashers in the past 12 months, on methods used to pay bills or receive income, or where the household indicated
that it did not pay bills or receive income in a typical month.
76
Households in underbanked group 1 were classified as underbanked because either they used credit AFS in the past 12 months, or they used transaction AFS in the
past 12 months but not to pay bills or receive income in a typical month.
77
61
Table 10.1 Selected Household Characteristics by Banking Status, 2017
For all households that paid bills and received income in a typical month, column percent
Characteristics
Underbanked
group 1
Underbanked
group 2
Fully
banked
Unbanked
Family income (Percent)
Less than $15,000
8.3
18.5
7.9
45.5
$15,000 to $30,000
12.3
22.9
12.8
30.7
$30,000 to $50,000
22.3
26.2
18.9
17.8
$50,000 to $75,000
22.5
17.3
19.7
3.7
At least $75,000
34.5
15.1
40.8
2.2
No high school diploma
7.9
16.1
6.2
33.4
High school diploma
23.6
31.6
23.7
36.5
Some college
30.3
34.4
28.6
24.0
College degree
38.2
17.9
41.5
6.1
15 to 24 years
6.1
10.2
4.2
8.0
25 to 34 years
18.5
21.8
15.0
23.1
35 to 44 years
19.0
20.3
15.6
19.8
45 to 54 years
19.9
17.2
17.8
19.6
55 to 64 years
17.9
18.0
19.4
17.3
65 years or more
18.6
12.5
28.0
12.2
Black
17.4
27.9
9.3
36.6
Hispanic
16.6
22.1
9.0
28.4
Asian
7.2
2.2
5.2
1.2
White
57.1
44.9
75.2
30.5
Other
1.8
2.9
1.3
3.3
Disabled, age 25 to 64
8.3
14.6
6.5
22.6
Not disabled, age 25 to 64
67.0
62.6
61.3
57.2
Not applicable (not age 25
to 64)
24.7
22.7
32.2
20.2
Income was about the same
each month
75.5
66.8
80.4
69.8
Income varied somewhat
from month to month
19.7
26.0
16.2
20.9
Income varied a lot from
month to month
4.7
7.1
3.3
9.3
Unknown
0.1
-
0.1
-
Education (Percent)
Age group (Percent)
Race/Ethnicity (Percent)
Disability status (Percent)
Monthly income volatility
(Percent)
Notes: Estimates may differ from those presented elsewhere in the report because of differences in the samples being analyzed. Households in underbanked group 1
used only bank methods to pay bills and receive income in a typical month, while households in underbanked group 2 did not exclusively use bank methods to pay
bills and receive income in a typical month. The - symbol indicates an estimate of zero. The population proportion may be slightly greater than zero. See Appendix
Table H.1 for additional socioeconomic and demographic characteristics of households in underbanked groups 1 and 2.
62 | 2017 FDIC National Survey of Unbanked and Underbanked Households
young, black, Hispanic, or working-age disabled; and were
more likely to have volatile income. For example, 18.5 percent
of households in underbanked group 2 had less than $15,000
in income, compared with 8.3 percent of households in underbanked group 1 and 7.9 percent of the fully banked. Likewise, 16.1 percent of households in underbanked group 2 did
not have a high school diploma, compared with 7.9 percent
of households in underbanked group 1 and 6.2 percent of the
fully banked.
Bank Account Access Methods
The share of households with volatile income was similar for
underbanked group 2 and the unbanked: 33.2 percent of
households in underbanked group 2 had income that varied
somewhat or a lot from month to month, compared with 30.2
percent of the unbanked.
As presented in Figure 10.2, almost all (97.9 percent) households in underbanked group 2 used at least one bank method
to pay bills or receive income in a typical month.79 Among
the other methods used, cash and transaction AFS were the
most prevalent (primarily for paying bills, as discussed in
section 9).80 Almost two-thirds of households in underbanked
group 2 (63.8 percent) used cash to pay bills or receive income in a typical month, a share that is almost as high as the
estimate for the unbanked (70.6 percent).
As illustrated in Figure 10.1, use of mobile banking as the
primary method of bank account access was similar across
the two underbanked groups.78 However, among households
in underbanked group 2, use of bank tellers was more prevalent and use of online banking less prevalent, compared with
households in underbanked group 1.
Methods Used to Pay Bills and Receive Income in a
Typical Month
Figure 10.1 Selected Primary Methods Used to Access Bank Accounts by Banking Status, 2017 (Percent)
39.8
36.0
25.9
23.8
22.4
20.2
18.7
17.7
14.9
Bank teller
Online banking
Underbanked group 1
Underbanked group 2
Mobile banking
Fully banked
Notes: Estimates may differ from those presented elsewhere in the report because of differences in the samples being analyzed. Households in underbanked group
1 used only bank methods to pay bills and receive income in a typical month, while households in underbanked group 2 did not exclusively use bank methods to pay
bills and receive income in a typical month.
The analysis of bank account access methods further excludes 323 observations (representing roughly 1.1 million banked households) that did not access their
accounts in the past 12 months or that did not report whether they accessed their accounts.
78
By definition, households in underbanked group 1 used only bank methods to pay bills and receive income in a typical month, so this group is not displayed in Figure
10.2.
79
Overall use of prepaid cards in the past 12 months was also higher among households in underbanked group 2 (21.7 percent), compared with households in
underbanked group 1 (7.5 percent).
80
63
Figure 10.2 Selected Joint Methods of Paying Bills and Receiving Income in a Typical Month by Banking Status, 2017
(Percent)
97.9
99.8
70.6
63.8
51.2
29.9
24.4
14.6
Any bank method
47.8
Cash
Underbanked group 2
13.7
1.9
0.0
Prepaid card
Transaction AFS
Fully banked
Unbanked
Notes: Estimates may differ from those presented elsewhere in the report because of differences in the samples being analyzed. Households in underbanked group
2 did not exclusively use bank methods to pay bills and receive income in a typical month. “Any bank method” includes households that used at least one bank
method to pay bills or receive income. “Transaction AFS” includes households that used a nonbank money order to pay bills or that used a nonbank check casher
to get the money from income received by paper check or money order.
Saving for Unexpected Expenses or Emergencies
As displayed in Figure 10.3, households in underbanked
group 1 saved for unexpected expenses or emergencies at
a similar rate to the fully banked, while households in underbanked group 2 saved at a somewhat lower rate.81
As indicated in Figure 10.4, among households that saved
for unexpected expenses or emergencies, the proportion that
kept savings in a savings account was similar for underbanked group 1 and the fully banked but lower for underbanked group 2. The proportion that kept savings in a savings
or checking account followed a similar pattern: 86.1 percent for underbanked group 1 and 89.2 percent for the fully
banked, compared with 74.0 percent for underbanked group
2. The pattern was opposite for maintaining savings in the
home, or with family or friends.
Figure 10.3 Rates of Saving for Unexpected Expenses or
Emergencies by Banking Status, 2017
62.9
61.8
52.3
21.8
Underbanked group 1
Underbanked group 2
Fully banked
Unbanked
Notes: Estimates may differ from those presented elsewhere in the report
because of differences in the samples being analyzed. Households in underbanked group 1 used only bank methods to pay bills and receive income in a
typical month, while households in underbanked group 2 did not exclusively
use bank methods to pay bills and receive income in a typical month.
The difference in savings rates between the two underbanked groups was no longer statistically significant after accounting for differences between the two groups
in the household characteristics listed in Appendix Table A.2. Almost all of the difference in savings rates between the two groups can be attributed to differences in
income and the other household characteristics between the two groups.
81
64 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Figure 10.4 Selected Savings Methods for Households That Saved by Banking Status, 2017 (Percent)
75.5
72.3
66.8
57.9
24.4 22.5 24.0
23.3
10.4
2.4
2.3
Savings account
Checking account
Underbanked group 1
11.2
7.8
0.1
In home, or with family or friends
Underbanked group 2
Fully banked
0.7
0.1
Prepaid card
Unbanked
Notes: Estimates may differ from those presented elsewhere in the report because of differences in the samples being analyzed. Households in underbanked
group 1 used only bank methods to pay bills and receive income in a typical month, while households in underbanked group 2 did not exclusively use bank
methods to pay bills and receive income in a typical month. Bars may sum to more than 100 percent because households were asked to select all savings
methods used.
Mainstream Credit Product Use and Credit
Characteristics
As illustrated in panel A of Table 10.2, use of most mainstream
credit products was lower among households in underbanked
group 2, relative to households in underbanked group 1. For
example, 48.9 percent of households in underbanked group 2
had a credit card from Visa, MasterCard, American Express,
or Discover, compared with 72.8 percent of households in
underbanked group 1. Further, the proportion of households
that did not have mainstream credit was much higher for
underbanked group 2 (28.8 percent) than for underbanked
group 1 (13.4 percent). Use of mainstream credit products for
underbanked group 1 was generally similar to that of the fully
banked.
Panel B of Table 10.2 shows some differences in credit
characteristics across the underbanked groups. Households
in underbanked group 2 were more likely than households in
underbanked group 1 to have felt discouraged about applying
for a credit card or bank personal loan or to have fallen
behind on bills. The proportion of households in underbanked
group 2 that fell behind on bills (38.1 percent) was similar to
the proportion among the unbanked (43.4 percent).82
Summary
Overall, the segmentation analysis suggests that it is
important to consider intensity of transaction AFS use in
measuring the underbanked. While some households make
incidental use of transaction AFS, other households use
transaction AFS to meet their basic financial needs, such as
paying bills or receiving income. If intensity of transaction
AFS use were considered in the classification of underbanked
households, fewer households in underbanked group 1 may
be classified as underbanked.
Differences between the two underbanked groups in the proportions of households that applied for or were denied a credit card or bank personal loan (or that were
denied, conditional on applying) were no longer statistically significant after accounting for differences between the two groups in the household characteristics listed
in Appendix Table A.2.
82
65
Table 10.2 Mainstream Credit Product Use and Credit Characteristics by Banking Status, 2017
For all households that paid bills and received income in a typical month, column percent
Underbanked
group 1
Underbanked
group 2
Fully
banked
Unbanked
Credit card
72.8
48.9
77.5
8.7
Store credit card
46.6
30.6
46.8
4.4
Mortgage, home equity loan,
or HELOC
34.5
19.0
39.4
4.0
Auto loan
42.1
30.0
34.5
7.0
Student loan
22.6
19.3
16.9
5.2
Bank personal loan
9.6
7.6
7.2
1.2
Other mainstream nonbank
3.8
5.0
1.6
1.3
No mainstream credit
13.4
28.8
12.9
77.1
Applied
19.9
16.9
14.4
3.1
Denied
5.6
6.9
2.0
2.0
Felt discouraged about
applying
10.2
16.6
3.3
9.6
Fell behind on bills
19.6
38.1
8.3
43.4
Denied, conditional on
applying
28.4
40.6
13.9
NA
A. Mainstream credit
product use (Percent)
B. Credit characteristics
(Percent)
Notes: Estimates may differ from those presented elsewhere in the report because of differences in the samples being analyzed. Households in underbanked group
1 used only bank methods to pay bills and receive income in a typical month, while households in underbanked group 2 did not exclusively use bank methods to
pay bills and receive income in a typical month. NA indicates that the sample size was too small to produce a precise estimate.
66 | 2017 FDIC National Survey of Unbanked and Underbanked Households
2017 FDIC National Survey of Unbanked and Underbanked
Households
11. Implications and Conclusions
The survey results show that the unbanked rate declined 0.5
percentage points between June 2015 and June 2017. This
decline can be attributed almost entirely to improvements
in the economic circumstances of U.S. households. The
unbanked rate fell for many groups that had high unbanked
rates in 2015. However, unbanked rates for these groups
remain substantially higher than the overall unbanked rate
in 2017. Below, we discuss a number of opportunities
to increase the use of mainstream banking services by
unbanked and underbanked households that may help to
further reduce unbanked and underbanked rates going
forward.
1. New underwriting technologies could help expand
access to small-dollar credit for banked consumers,
including consumers with little or no credit history.
The vast majority of the 13 percent of households
with unmet demand for mainstream small-dollar
credit are banked, and almost all receive income and
pay bills using their bank accounts. But few of these
households applied for a credit card or bank personal loan. Account balances and transactions may provide information for banks to underwrite small-dollar
credit to some of these households.
holds with unmet demand for mainstream small-dollar credit
applied for a credit card or bank personal loan in the past 12
months.
Among households with unmet demand for mainstream
small-dollar credit, almost six in ten indicate that they were
current on their bills over the past 12 months. While this is
an incomplete measure of creditworthiness, it nevertheless
provides some insight into the financial situation of these
households. For some of the remaining households, it is possible that obtaining mainstream small-dollar credit could have
prevented the household from falling behind on its bills.
For households with a banking relationship, account balances
and transactions may provide information for underwriting
loans to these consumers. In particular, such information may
enable banks to underwrite small-dollar loans to consumers
with little or no credit history. Almost a quarter of banked
households with unmet demand for mainstream small-dollar
credit likely have insufficient credit history to have a credit
score.83 Providing small-dollar bank loans to these households may help strengthen their relationships with banks and
allow them to begin building credit.84
Access to small-dollar credit is important for weathering
financial setbacks, particularly for households with fluctuating
income or lack of savings. Almost 13 percent of households
have unmet demand for mainstream small-dollar credit,
meaning that, in the past 12 months, they applied for and
were denied a credit card or bank personal loan, felt discouraged about applying for a credit card or bank personal loan,
or used credit AFS.
2. A
bout one in five households likely have little or no
credit history. The vast majority of these households
are banked and may not seek credit until a need
arises. Helping these households establish and build
a credit history may particularly benefit black households, Hispanic households, and households headed
by a working-age individual with a disability. All of
these households are disproportionately less likely to
have mainstream credit.
Nine in ten of these households are banked, and more than
eight in ten have been banked for 12 months or longer. Most
direct deposit their income into their bank accounts (87.1
percent) and pay bills using methods directly linked to their
accounts, including electronic payment from a bank account,
bank debit card, or personal check (95.0 percent). Despite an
active banking relationship, only one-third of banked house-
Approximately one in five households did not, in the past
12 months, have any mainstream credit products that are
reported to credit bureaus and therefore likely have little or no
credit history. Although three-quarters of these households
are banked, they might not be aware of the importance of
credit and might not seek credit until a need arises.
In the past 12 months, these households did not have any of the credit products that are likely reported to credit bureaus. These include credit cards; store credit
cards; mortgages, home equity loans, and home equity lines of credit (HELOCs); auto loans; student loans; bank personal loans; and other mainstream nonbank credit.
83
In qualitative research, several banks described small-dollar loan products that they offered, some of which included financial education. See “Bank Efforts to Serve
Unbanked and Underbanked Consumers Qualitative Research,” May 25, 2016 (available at http://www.fdic.gov/consumers/community/research/qualitativeresearch_
may2016.pdf).
84
67
Some segments of the population are less likely to have
had mainstream credit in the past 12 months. More than
one-third (36.0 percent) of black households and 31.5
percent of Hispanic households had no mainstream credit,
compared with 14.4 percent of white households. Four in
ten working-age disabled households (40.4 percent) had no
mainstream credit, compared with 15.3 percent of workingage nondisabled households.
These disparities persist even after accounting for
socioeconomic and demographic characteristics (such as
income, education, and age) and bank account ownership.
For example, among households with income between
$30,000 and $50,000, 27.9 percent of black households and
28.5 percent of Hispanic households had no mainstream
credit, compared with 16.2 percent of white households.
Helping households with no mainstream credit establish and
build a credit history may therefore particularly benefit black,
Hispanic, and working-age disabled households that are
disproportionately less likely to have mainstream credit.
3. M
obile banking holds real promise for deepening
the connection between underbanked households
and their banks while increasing the safety and
convenience of bill payments. A large share of
underbanked households pays bills in a typical
month with cash or nonbank money orders. More
than two in five of these households already use
mobile banking to access their bank accounts.
Increased use of mobile banking activities by these
households may enable them to conduct a greater
share of their basic financial transactions within the
banking system.
About two in five underbanked households pay some bills in
a typical month with cash or nonbank money orders. These
underbanked households demonstrate a high level of engagement with the banking system: more than four in five
also pay bills in a typical month using a bank method, such as
electronic payment from a bank account. A similar proportion
typically receives income through direct deposit into their
bank accounts.
These underbanked households’ high level of engagement
with the banking system suggests that they may be receptive
to conducting a greater share of their basic financial transactions within the banking system. Banks and other stakeholders could encourage and facilitate the use of mobile bill pay
or mobile person-to-person payments by these households
because more than four in five of them had access to a
smartphone.
Also, more than two in five of these underbanked households
already use mobile banking to access their accounts. But only
a quarter use a bank’s mobile website or mobile app to pay
bills, and only about one in eight use a bank’s mobile website
or mobile app to send money to other people. Using mobile
bill pay or mobile person-to-person payments instead of cash
or nonbank money orders increases safety and convenience,
deepens the connection between households and their
banks, and increases the opportunity for households to derive
value from the banking relationship.
To the extent that the use of cash or money orders is partially
the result of payee requirements, efforts to encourage and
make it easier for a range of payees (e.g., landlords) to accept
electronic payments may help these households reduce their
use of cash and nonbank money orders. For example, opportunities may exist for banks to increase customer awareness
about innovations that have made mobile payments between
individuals and payments to businesses faster and safer.
4. P
hysical access to bank branches remains important
even as use of mobile banking and online banking
has increased. In 2017, the great majority of banked
households visited a bank branch in the past 12
months, and more than one-third visited ten or more
times. In addition, almost one in six unbanked households visited a bank branch in the past 12 months.
These findings suggest that branches continue to
play an important role for banked households and
that opportunities may exist for branch staff to
inform unbanked households about products and
services that can help meet their financial needs.
Household use of mobile banking as the primary method
of bank account access more than doubled between 2013
and 2017. Use of online banking as the primary method also
increased during that period. Commensurately, the proportion
of households that primarily use bank tellers to access their
accounts declined from 32.2 percent in 2013 to 24.3 percent
in 2017. Despite this decline, physical access to branches
remains important.
In 2017, almost three in four banked households used bank
tellers to access their accounts at least once in the past 12
months, a higher proportion than any other method asked
about in the survey. Moreover, some households may rely on
bank branches for activities other than accessing an account,
such as resolving a problem or asking about products or
services. Almost five in six banked households visited a bank
branch at least once in the past 12 months, and more than
one in three visited ten or more times.
68 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Branch visits were prevalent even among banked households
that did not use bank tellers as their primary method of
account access. For example, about eight in ten banked
households that primarily used mobile banking visited a
branch in the past 12 months, and nearly one-quarter visited
ten or more times. These findings suggest that branches
and the range of services they provide continue to play an
important role for many banked households.
In addition, some groups with higher unbanked and underbanked rates, including lower-income households, less-educated households, older households, and households in rural
areas, continue to disproportionately use bank tellers as their
primary—and often only—method for accessing their accounts. For example, in 2017, more than one-third of banked
households in rural areas primarily used bank tellers to access
their accounts, and one in five only used bank tellers. Almost
half of banked households in rural areas visited a branch ten
or more times in the past 12 months.
While bank branch visits were less common among unbanked
households, relative to banked households, almost one in six
unbanked households visited a bank branch at least once in
the past 12 months, and one in twenty visited ten or more
times. Approximately seven in ten unbanked households that
visited a branch had previously been banked, and about two
in five were “very likely” or “somewhat likely” to open a bank
account in the next 12 months.
These findings suggest that unbanked households that visit
banks find value in the banking system and can benefit from
the development of products and services that could meet
their financial needs.85 For example, if unbanked households
are visiting a bank branch to cash a payroll check, banks that
offer accounts with low or no minimum balance requirements
and low fees could promote these accounts and advertise the
convenience and security of using direct deposit and bill pay
compared with cashing a payroll check.86
5. U
nbanked rates for some segments of the population
have declined as economic conditions improved between 2011 and 2017. Still, unbanked rates for these
groups, including black and Hispanic households,
remain substantially above the national average. At
the same time, unbanked rates for other population
segments, such as working-age disabled households, have remained high and stayed fairly constant
between 2011 and 2017. Understanding the evolution
of unbanked rates for different population segments
and adopting targeted strategies may help sustain
increases in bank account ownership in future economic downturns and increase access for different
population segments with high unbanked rates.
Almost 17 percent (16.9 percent) of black households and
14.0 percent of Hispanic households were unbanked in 2017.
Unbanked rates for these two groups have declined steadily in recent years, consistent with the overall decline in the
national unbanked rate. For black households, the unbanked
rate has fallen from 21.4 percent in 2011. Similarly, the
unbanked rate for Hispanic households has decreased from
20.1 percent in 2011.
Improved economic conditions of black households in 2017
relative to 2011, particularly increases in household income,
educational attainment, and employment, explain almost
all of the 4.5 percentage point decline in the unbanked rate
for black households over this period.87 Improved economic
conditions of Hispanic households in 2017 relative to 2011
also played a role in the declining unbanked rate for those
households.88
These findings are consistent with findings from the 2013 survey, where job losses or gains and significant income changes
seemed to be common triggers for bank account openings
and closings among households that had recently become
banked or had recently become unbanked.89 To reduce the
likelihood that future economic downturns reverse some or all
See “Bank Efforts to Serve Unbanked and Underbanked Consumers Qualitative Research,” May 25, 2016, for examples of a range of products and services that the
11 interviewed banks offered to sustainably meet the needs of unbanked and underbanked consumers. This report also describes a variety of additional strategies
used by these banks to reach and serve unbanked and underbanked consumers.
85
86
See http://www.fdic.gov/consumers/template/template.pdf for the FDIC Model Safe Accounts Template.
A linear probability model was estimated to account for changes from 2011 to 2017 in the distribution of households across the household characteristics listed
in Appendix Table A.2 (except for monthly income volatility, which is not available for 2011). Changes in these household characteristics between 2011 and 2017
accounted for about 85 percent of the decrease in the unbanked rate for black households over this period.
87
A linear probability model was estimated to account for changes from 2011 to 2017 in the distribution of households across the household characteristics listed
in Appendix Table A.2 (except for monthly income volatility, which is not available for 2011). Changes in these household characteristics between 2011 and 2017
accounted for approximately 40 percent of the decrease in the unbanked rate for Hispanic households over this period.
88
See 2013 FDIC National Survey of Unbanked and Underbanked Households, October 2014 (available at http://www.economicinclusion.gov/surveys/2013household/
documents/2013_FDIC_Unbanked_HH_Survey_Report.pdf).
89
69
of the decline in the unbanked rate for population segments
whose unbanked rates have declined, policymakers and industry participants may consider ways to cushion the impact
of adverse financial shocks on a household’s ability or desire
to maintain a bank account, such as forbearance of fees or
the use of flexible product design.
Moreover, even with the declines in unbanked rates, bank
account ownership among black and Hispanic households
continues to be significantly below the national average. Further, an overwhelming majority of unbanked black, Hispanic,
and working-age disabled households are “not very likely” or
“not at all likely” to open an account in the next 12 months.
While unbanked rates have declined for black and Hispanic
households during this period of economic expansion, unbanked rates for other populations with a large percentage of
unbanked households have not declined at a similar pace. For
example, the unbanked rate for working-age disabled households has been fairly constant between 2011 and 2017: 18.9
percent in 2011, 18.4 percent in 2013, 17.6 percent in 2015,
and 18.1 percent in 2017.
These findings suggest the continued need for targeted
strategies that address barriers to bank account ownership for each of the different population segments with high
unbanked rates.90 A substantial portion of the 6.2 percentage
point decline in the unbanked rate for Hispanic households
between 2011 and 2017 cannot be explained by changes in
economic conditions. Research to identify the factors that
have contributed to this decline, beyond those related to the
business cycle, can inform actions that may further reduce
the unbanked rate for Hispanic households, and these efforts
may be adaptable to other groups with high unbanked rates.
See “Bank Efforts to Serve Unbanked and Underbanked Consumers Qualitative Research,” May 25, 2016, for examples of different efforts and targeted strategies
undertaken by 11 interviewed banks to reach and serve unbanked and underbanked consumers.
90
70 | 2017 FDIC National Survey of Unbanked and Underbanked Households
2017 FDIC National Survey of Unbanked and Underbanked
Households
Appendix 1. FDIC Technical Notes
The data for this report were collected through a Federal
Deposit Insurance Corporation (FDIC)-sponsored Unbanked/
Underbanked Supplement to the Current Population Survey (CPS) for June 2017. The CPS is a monthly survey of
about 52,000 interviewed households conducted by the U.S.
Census Bureau for the Bureau of Labor Statistics (BLS). The
survey is based on a scientific sample that is representative of
the U.S. civilian, noninstitutionalized population, aged 15
or older.
The CPS is the primary source of information on the labor
force characteristics of the U.S. population, including employment, unemployment, and earnings statistics. The CPS
includes a variety of demographic characteristics, such as
age, sex, race, marital status, and educational attainment. Additional information about the CPS is provided on the Census
Bureau’s website.1
The CPS sample consists of independent samples in each
state and the District of Columbia. The sample sizes for each
state are set so that specific precision requirements for estimating unemployment rates will be met.2 The sample design
ensures that most of the households in a given state have the
same probability of being selected, though, in general, household selection probabilities will vary across states. Because
the CPS design is state-based, most of the estimates for the
Unbanked/Underbanked Supplement should be precise at the
state level and for some metropolitan statistical areas (MSAs).
Unbanked/Underbanked Supplement
The fifth Unbanked/Underbanked Supplement was conducted
in June 2017. The first, second, third, and fourth supplements
were conducted in January 2009, June 2011, June 2013, and
June 2015, respectively. The primary purpose of the supplement is to estimate the percentage of U.S. households that
are “unbanked” and “underbanked” and to identify the
reasons why. The supplement survey instrument used in
2017, attached as Appendix 3, included approximately 50
questions designed to provide this information.
The 2017 survey instrument is similar to previous survey
instruments. The 2009 instrument was developed with the
expertise of a national consulting firm, which specializes in
public opinion research, as well as input from the Census Bureau’s Demographic Surveys Division and the BLS. The 2009
survey instrument underwent four rounds of cognitive field
pre-testing and was revised to address the feedback gathered
from each round.3 The questionnaire was revised in 2011,
2013, 2015, and 2017. For a detailed description of the most
recent revisions, which underwent two rounds of cognitive
testing, see Appendix 2. Because of changes in the questionnaire, direct comparisons between 2017 and prior-year
estimates are not possible in some cases.
Eligibility and Exclusions
All households that participated in the June 2017 CPS were
eligible to participate in the Unbanked/Underbanked Supplement. However, only households whose respondents
specified that they had some level of participation in their
household finances and responded “Yes” or “No” to whether someone in their household had a bank account (survey
supplement Question 2, or Q2) were considered survey respondents.4 CPS household respondents who did not answer
or answered “Don’t know” to Q2, or who did not participate
in their household financial decisions (or refused to answer),
were asked no further questions and were classified as nonrespondents for the supplement.
Coverage and Response Rates
For the June 2017 CPS, a statistical sample of 60,843
survey-eligible households was selected from the sampling
frame.5 Of these households, 52,068 participated in the CPS,
See, for example, U.S. Census Bureau’s Technical Paper 66, “Design and Methodology, Current Population Survey,” available at http://www.census.gov/
prod/2006pubs/tp-66.pdf.
1
The precision targets that are the basis for the sample design of the CPS are provided in Chapter 3 of Technical Paper 66, available at http://www.census.gov/
prod/2006pubs/tp-66.pdf.
2
The goal of each round was to determine respondents’ comprehension of each question, test the flow of the questions, find major recall difficulties, ascertain the
sensitivity or inappropriateness of any questions, and gauge the operational feasibility of the supplement. No changes to the survey were recommended following the
fourth round of testing.
3
Respondents involved in their household finances include respondents in households where adults have separate finances or in households where the respondent
was the only adult in the household. For households where adults share finances or have a mix of shared and separate finances, respondents were asked to specify
how much they participated in their household financial decisions. Only those who reported having at least some level of participation were considered to be involved
in their household finances.
4
For details on the sampling frame, refer to the technical documentation for the June 2017 supplement, available at http://thedataweb.rm.census.gov/ftp/cps_ftp.html.
5
71
resulting in an 86 percent response rate. There were 8,775
nonrespondent eligible households. Most of these nonrespondents either refused to participate (76 percent) or were not
home at the time of the interview visit or call (13 percent). The
remaining 11 percent consisted of households where (a) the
household respondent was temporarily absent, (b) the household could not be located, (c) language barriers prevented the
interview, or (d) other reasons. Because of the availability of
translators for many languages, only 0.5 percent of the nonrespondents (44 households) did not participate as a result of
language barriers.
Coverage ratios for the CPS are derived as a measure of the
percentage of persons in the target universe (the U.S. civilian,
noninstitutionalized population, aged 15 or older) that are
included in the sampling frame.6 The overall coverage ratio for
the June 2017 CPS was 89 percent. The missing 11 percent
consists of three groups: (1) persons residing in households
that are not in the CPS sampling frame, (2) noninstitutionalized persons not residing in households at the time the CPS
was conducted, and (3) household residents who were not
listed as household members for the CPS for various reasons.
The coverage ratios varied across demographic groups. For
example, among women aged 15 and older, the coverage
ratio was 93 percent for whites, 80 percent for blacks, and 84
percent for Hispanics.
Of the 52,068 households that participated in the CPS, 35,217
(68 percent) also participated in the Unbanked/Underbanked
Supplement.7 Supplement survey response rates vary by
household characteristics, ranging from 60 to 73 percent for
the segments of the population listed in Appendix Table A.2.
The weights calculated by the Census Bureau for the CPS
and the Unbanked/Underbanked Supplement respondents
were adjusted to account for both nonresponse and undercoverage. These weight adjustments help correct any biases
in estimates because of nonresponse and undercoverage, so
that results from the CPS are representative of the U.S. civilian, noninstitutionalized population, aged 15 or older.8
Analysis of Supplement Survey Results
Using supplement survey results, households were classified
as unbanked if they answered “No” to the question, “Do you
or anyone else in your household have a checking or savings account now?” Households that answered “Yes” to this
question were classified as underbanked if they indicated that
they used one of the following products or services from an
alternative financial services provider in the past 12 months:
money orders, check cashing, international remittances,
payday loans, refund anticipation loans, rent-to-own services,
pawn shop loans, or auto title loans.
The estimated proportion of U.S. households that are unbanked was derived by dividing the sum of the weights of
the household respondents who were identified as being
unbanked by the sum of the weights of all household respondents. The same formula was used to estimate the proportion
of U.S. households that are underbanked. For estimated
proportions of unbanked or underbanked households for demographic subgroups, the same computational approach was
used and applied to respondent households in the subgroup.
In addition to presenting estimated proportions, many of the
tables in this report include estimated numbers of households
(e.g., total households, unbanked households, or underbanked households). An estimated number of households for
a given category, such as unbanked, is derived as the sum
of the weights of the sample households in that category.
For example, for the entire supplement sample of 35,217
respondent households, the sum of the household weights is
roughly 129.3 million, which would be an estimate of all U.S.
households as of June 2017. The Housing Vacancy Survey, another survey related to the CPS that uses household
controls to produce household weights, provided an estimate
of 119.1 million as the number of households in June 2017.9
This difference (129.3 million versus 119.1 million) is because
household weights prepared by Census for the CPS and for
this supplement survey are generally taken to be the reference
person weights and are not adjusted to align with household
count controls. Household count controls were not used to
adjust household weights because the CPS is a person-level survey rather than a household-level survey; therefore,
universe controls were used only in the preparation of person
weights. As a result, the sum of household weights shown in
our tables for a category tends to be somewhat higher than
the actual household count for the category.
The coverage ratio is the weighted number of persons in a demographic group (after weights are adjusted to account for household nonresponse) divided by an
independent count of persons in that demographic group (obtained from the 2010 Census with updates based on the American Community Survey).
6
Taking into account the nonresponse to the base CPS, the overall response rate for the Unbanked/Underbanked Supplement was 58 percent.
7
This adjustment is done by introducing three stages of ratio estimation that adjust weights to align with population control totals (independent population estimates
for various demographic and geographic groups). The household weight is generally taken to be the weight of the householder/reference person; however, if the
householder/reference person is a married male, the spouse’s weight is used.
8
See Table 13a Monthly Household Estimates: 2000 to Present, Vintage 2017 (July 26, 2018), available at http://www.census.gov/housing/hvs/data/hist_tab_13a_
v2017.xlsx.
9
72 | 2017 FDIC National Survey of Unbanked and Underbanked Households
This report also contains a number of tables for which unbanked percentages and other household statistics are computed for subgroups defined by a particular socioeconomic
or demographic characteristic. The household classification
of a socioeconomic or demographic variable that is defined
at the person level rather than the household level (e.g., race/
ethnicity, education, or employment status) is based on the
socioeconomic or demographic classification of the householder/reference person (i.e., the person who owns or rents
the home).10
The Census Bureau classifies households into different
household types. For instance, a family household is a
household that includes two or more people related by birth,
marriage, or adoption and residing together, along with any
unrelated people who may be residing there. Detailed definitions regarding household types can be found in the technical
documentation on the CPS website.11
Households are categorized into racial/ethnic classifications as follows: if the householder is identified as black,
the household is classified as “black” regardless of whether
the householder is identified as Hispanic or any other race.
If the householder is not identified as black and is identified
as Hispanic, the household is classified as “Hispanic.” If the
householder is identified as Asian and not black or Hispanic,
then the household is classified as “Asian.” If the householder
is identified as white and not any other race and not Hispanic, then the household is classified as “white.” All remaining
households are classified as “other.”
This report provides unbanked and other estimates for the
population of households with disabilities. As in the 2013
report (the first time these estimates were presented) and
the 2015 report, households are categorized as follows: if
the householder is between age 25 and 64 and either (a)
indicates “Yes” to any of the six-question disability sequence
in the base CPS or (b) is classified as “Not in labor force –
disabled,” the household is classified as “Disabled, age 25
to 64.”12 If the householder is between age 25 and 64 and
neither condition (a) nor (b) above is met, the household is
classified as “Not disabled, age 25 to 64.” If the householder
is not between the ages of 25 and 64, the household is classified as “Not applicable (not age 25 to 64).”13
This report presents estimates of unbanked and underbanked
rates (and other outcomes of interest) for larger metropolitan
statistical areas (MSAs). MSA delineations are established by
the Office of Management and Budget (OMB). OMB published
a revised set of MSA delineations in February 2013, based
on data from the 2010 Census and the 2006-2010 American
Community Surveys. The 2013 delineations superseded the
earlier delineations based on Census 2000 data, first established by OMB in June 2003.14
As discussed in the technical documentation to the June
2015 supplement, the Census Bureau phased the 2013 MSA
delineations into the CPS (and phased out the 2003 delineations) over the period May 2014 to July 2015.15 Housing units
first included in the CPS before May 2014 were assigned
metropolitan area codes based on the 2003 delineations.
These metropolitan area codes consisted of metropolitan New
England city and town area (NECTA) codes for New England
states (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont) and MSA codes for other
states.16 Housing units first included in the CPS in May 2014
or later were assigned metropolitan area codes based on the
2013 delineations. These metropolitan area codes consisted
only of MSA codes, as housing units in New England were
given MSA codes as part of the phase-in of the 2013
delineations.
For the 2017 survey data, all housing units were assigned
metropolitan area codes based on the 2013 delineations. For
the 2015 survey data, approximately three-quarters of housing units were assigned metropolitan area codes based on
the 2013 delineations, while the remaining housing units were
assigned metropolitan area codes based on the 2003 delinea-
In a few cases, the householder/reference person is classified as an ineligible respondent for the CPS, but another eligible household resident participated in the CPS
and in the Unbanked/Underbanked Supplement. In these cases we use the attributes of the eligible respondent to characterize the household.
10
11
See http://www.census.gov/programs-surveys/cps/technical-documentation/subject-definitions.html.
Specifically, we use the variable PEMLR (monthly labor force recode) to determine if the respondent is not in the labor force because of a disability. Refer to the CPS
Data Dictionary for detail on the six-question disability sequence, available at the following link: http://thedataweb.rm.census.gov/ftp/cps_ftp.html.
12
A universally accepted method to identify the population with disabilities does not exist. Key estimates from the FDIC Unbanked/Underbanked Supplement, such as
the proportion of disabled households that are unbanked, are qualitatively similar using alternative disability measures. See Appendix I of the 2013 report for details,
available at http://www.economicinclusion.gov/surveys/2013household/documents/2013_FDIC_Unbanked_HH_Survey_Appendix.pdf.
13
For February 2013 delineations, see OMB Bulletin Number 13-01 (February 28, 2013), available at http://www.whitehouse.gov/sites/whitehouse.gov/files/omb/
bulletins/2013/b13-01.pdf. For June 2003 delineations, see OMB Bulletin Number 03-04 (June 6, 2003), available at http://www.whitehouse.gov/wp-content/
uploads/2017/11/bulletins_b03-04.pdf. In each year between 2003 and 2009, OMB published minor revisions to the MSA delineations, based on the Census Bureau’s
annual population estimates.
14
15
The technical documentation for the June 2015 supplement is available at http://thedataweb.rm.census.gov/ftp/cps_ftp.html.
Unlike MSAs, which are made up of one of more full counties or county equivalents, NECTAs are composed of cities and towns and often do not follow county
boundaries.
16
73
tions. To facilitate MSA-level estimates using the 2015 survey
data, a housing unit with an obsolete 2003 MSA code was
assigned the corresponding 2013 MSA code.17 A housing unit
sample replicates (and on the CPS methodology in general)
are available from the Census Bureau.19
with a NECTA code was assigned the 2013 MSA code that
comprised the majority of the NECTA population.18 Overall,
less than 3 percent of housing units in the 2015 survey data
were affected by these adjustments.
Estimated differences discussed in this report are significant
at the 10 percent level, unless noted otherwise. That is, if
the population difference were zero, then the probability of
obtaining estimates having the observed difference or a larger
difference would be no more than 10 percent and could be
considerably less. For example, the estimated difference
in the proportions of U.S. households that were unbanked
between 2017 (6.5 percent) and 2015 (7.0 percent) is -0.5
percentage points. The estimated standard error of this difference (computed using the 160 replicates as described above)
is 0.2 percentage points. Under the assumption that the true
difference in the unbanked rate between 2017 and 2015 is
zero, the probability of observing the -0.5 percentage point
difference in our sample data is 3.7 percent (i.e., the p-value
is 0.037).
For the 2013 and earlier survey data, all housing units were
assigned metropolitan area codes based on the 2003 delineations. For these survey years, metropolitan area estimates
provided in this report are based on the 2003 delineations.
Because of changes in geographic boundaries (e.g., the
addition or subtraction of a county), some metropolitan area
estimates that use 2017 and 2015 survey data are not directly
comparable to the corresponding metropolitan area estimates
that use 2013 and earlier survey data. In the report tables,
a tilde (~) next to an MSA name indicates that the MSA was
affected by a geographic boundary change. All MSA names in
the tables, however, reflect the 2013 delineations.
Statistical Precision of Estimates
To indicate the precision of certain estimates, standard errors
were calculated based on the variation of the estimates
across a set of 160 sample replicates provided by the Census
Bureau. Details of the calculation of standard errors based on
Certain 2017 report appendix tables include 90 percent confidence intervals in addition to point estimates. The confidence
interval is one way to describe the uncertainty surrounding the
estimate. For example, as shown in Appendix Table A.3, the
estimated proportion of U.S. households that were unbanked
in 2017 is 6.5 percent, and the 90 percent confidence interval
around this estimate ranges from 6.2 to 6.8 percent.
In the 2015 survey data, some housing units were located in counties populous enough to be identified, but no MSA code was assigned because these counties
were not in an MSA based on the 2003 delineations (all of these housing units were first included in the CPS before May 2014). Because some of these counties were
in an MSA based on the 2013 delineations, a 2013 MSA code was assigned to housing units located in such counties.
17
For example, housing units with a NECTA code for Boston-Cambridge-Quincy, MA-NH, were assigned the MSA code for Boston-Cambridge-Newton, MA-NH. For
each NECTA code in the 2015 survey data, at least 80 percent of the Census 2010 NECTA population (and the estimated July 1, 2015, NECTA population) resided
within the corresponding MSA, and for the majority of the NECTAs this number was at least 90 percent.
18
For a detailed description of the methodology used to calculate standard errors based on sample replicates, see Chapter 14 of Technical Paper 66, available at
http://www.census.gov/prod/2006pubs/tp-66.pdf.
19
74 | 2017 FDIC National Survey of Unbanked and Underbanked Households
2017 FDIC National Survey of Unbanked and Underbanked
Households
Appendix 2. 2017 Revisions to the FDIC National Survey of Unbanked and Underbanked
Households
The 2017 survey instrument is largely similar to the 2015
survey instrument. However, some revisions were made from
2015 to 2017 based on lessons learned from past survey
experience, cognitive testing of the 2017 instrument, and an
interest in certain economic inclusion topics not covered in
the 2015 instrument. In particular, the 2017 survey added new
questions about households’ visits to bank branches, use of
a mobile phone for banking activities, and use of mainstream
credit products.
To accommodate the new questions in the 2017 survey
instrument and satisfy space constraints, some questions
from the 2015 survey instrument were dropped. For example,
the 2017 survey did not include questions on households’
perceptions about banks’ interest in serving households like
theirs or on households’ learning about finances.
In 2017, banked households and recently unbanked households (i.e., households that did not have an account at the
time of the survey but did at some point in the 12 months
before) were asked whether they used a mobile phone to
check email from a bank about an account (Q80a) or whether
they received a mobile text alert or push notification from a
bank about an account (Q80b). Banked households that used
mobile banking to access an account in the past 12 months
were further asked whether they used a bank’s mobile website or bank’s mobile app to check a bank account balance
or recent transactions (Q80c), to make a bill payment (Q80d),
to send money to other people (Q80e), or to transfer money
between bank accounts owned by the same person (Q80f).
Banked households that used mobile banking to access an
account in the past 12 months were also asked whether they
used a mobile phone’s camera to deposit a check into a bank
account (Q80g).
Specific revisions to the 2017 survey are described below.
Bank Branch Visits Among Banked and Unbanked
Households
The 2017 survey included new questions about bank branch
visits. First, households that did not previously indicate
that they visited a bank branch in the past 12 months (i.e.,
unbanked households, or banked households that did not
access an account using a bank teller in the past 12 months)
were asked whether they spoke with a teller or other employee in person at a bank branch in the past 12 months (Q70).
Second, households that visited a bank branch in the past
12 months (i.e., households that answered “Yes” to Q70 or
that accessed an account using a bank teller in the past 12
months) were asked how many times they spoke with a teller
or other employee in person at a bank branch: one to four
times in the past 12 months, five to nine times in the past 12
months, or ten or more times in the past 12 months (Q71).
Mobile Activities
The 2017 survey included a series of questions on use of a
mobile phone for banking activities in the past 12 months.
These questions were not asked in 2015, although several of
these questions were asked in 2013.
All of the activities asked about in the 2017 survey, except for
whether households used a mobile phone to check email from
a bank about an account, were also asked in the 2013 survey
(2013 survey Q2i). The 2013 survey included some activities
not asked about in the 2017 survey, specifically, whether
households downloaded or used a bank’s mobile app, used
a mobile phone to locate the closest in-network ATM or bank
branch, or used a mobile phone for other activities.1
Prepaid Cards
The introductory language for the questions on prepaid card
use was changed slightly. The statement “I am not asking
about gift cards or debit cards linked to a checking account”
was moved to the end of the introductory paragraph. In 2015,
this statement was at the beginning of the introductory
paragraph.
Alternative Financial Services
A new question was added to gather information on households’ use of loans or lines of credit from alternative financial
services (AFS) providers that might not have been included
in their answers to preceding questions. Specifically, households that indicated that they did not have a payday loan
(Q122), pawn shop loan (Q123), refund anticipation loan
In the 2013 survey, all mobile activities were asked only of banked households that used mobile banking to access a bank account in the past 12 months. The
proportion of all banked households that received a mobile text alert or push notification from a bank about an account is therefore not comparable over time because
different types of households in the 2013 and 2017 surveys were asked about this activity.
1
75
(Q124), or auto title loan (Q126) in the past 12 months were
asked whether they had taken out any other types of loans
or lines of credit from a payday lender, auto title lender, pawn
shop, or check casher in the past 12 months (Q127). To
accommodate this new question, questions about auto title
loans (Q126) and rent-to-own services (Q125) were reordered.
Additionally, questions about sending money abroad in a
typical month were dropped (2015 survey Q132 and Q134),
and questions about places used to send money abroad in
the past 12 months were streamlined (Q131 and Q133 from
the 2015 survey were dropped and replaced with Q135 in the
2017 survey).
Mainstream Credit
The 2017 survey included a series of questions about use of
mainstream credit products, expanding on questions asked in
the 2015 survey. Specifically, households in the 2017 survey
were asked whether, in the past 12 months, they had a credit
card from Visa, MasterCard, American Express, or Discover
(Q1600a); a store credit card that could only be used at that
store (Q1600b); an auto loan (Q1600c); a mortgage, home
equity loan, or home equity line of credit (Q1600d); a student
loan (Q1600e); other personal loans or lines of credit from a
bank (Q1600f); or other personal loans or lines of credit from
a company other than a bank (Q1600g).2
The 2017 survey questions about credit cards from Visa,
MasterCard, American Express, or Discover and about
personal loans or lines of credit from a bank (Q1600a and
Q1600f) replaced similar questions from the 2015 survey
(2015 survey Q160 and Q161). The wording and location of
these questions changed somewhat to accommodate the
new questions in the 2017 survey about other types of
mainstream credit products.
First, in the introductory language to the questions about income receipt, “retirement” income was added to the example
list of income sources. Second, a question that asked households to choose the primary (i.e., most common) method of
bill payment in a typical month was dropped (2015 survey
Q151), and the remaining questions on income receipt and
bill payment methods were streamlined. Third, households
that did not indicate that they received income by any of the
methods asked about in the survey (Q140a-e) and that did not
volunteer that they did not receive income were asked a new
question about whether they received any income from work,
retirement, government benefits, or other sources in the past
12 months (Q140x). Similarly, households that did not indicate
that they paid bills by any of the methods asked about in the
survey (Q150a-i) and that did not volunteer that they did not
pay bills were asked a new question about whether they paid
any bills for things like mortgage, rent, utilities, or child care in
the past 12 months (Q150x).
Households’ Perceptions About Banks and
Households’ Learning About Finances
A question in the 2015 survey that asked households how
interested banks are in serving households like theirs (2015
survey Q101) was dropped. Questions about whether households sought financial information from banks in the past 12
months (2015 survey Q182) and about whether households
attended financial education classes or financial counseling
sessions in the past 12 months (2015 survey Q183 and Q184)
were also dropped.
Income Receipt and Bill Payment in a Typical Month
The 2015 survey included new questions about the ways
households receive income and pay bills in a typical month.
These questions were retained in the 2017 survey, with some
minor revisions.
For Q1600c, households that previously indicated that they had taken out an auto title loan were told that an auto loan is different from an auto title loan. For Q1600g,
households that previously indicated that they had taken out a payday loan, pawn shop loan, refund anticipation loan, auto title loan, or other types of loans or lines of
credit from a payday lender, auto title lender, pawn shop, or check casher were told to not include such loans when answering Q1600g.
2
76 | 2017 FDIC National Survey of Unbanked and Underbanked Households
2017 FDIC National Survey of Unbanked and Underbanked
Households
Appendix 3. Survey Instrument
Next, I’d like to ask you some questions about household finances.
1. Which of the following best describes your household finances? Do the adults…
¨ Share all finances
[CONTINUE]
¨ Share some finances
[CONTINUE]
¨ Share no finances at all
[SKIP TO Q2]
¨ I AM THE ONLY ADULT IN THE HOUSEHOLD (VOLUNTEERED)
[SKIP TO Q2]
¨ DK/REFUSE [CONTINUE]
1a. How much do you participate in making financial decisions for your household?
¨ A lot
[CONTINUE]
¨ Some [CONTINUE]
¨ Not at all
[TERMINATE]
¨ DK/REFUSE [TERMINATE]
2. Do you (if OTHERS AGE≥15 FILL: or anyone else in your household) have a checking or savings account now?
¨ YES [CONTINUE]
¨ NO
[SKIP TO Q3]
¨ DK/REFUSE [TERMINATE]
[Questions 2a-2h are asked only of households that have a bank account.]
2a. Who is that? (Enter Line Number)
¨ 1-16 [CONTINUE]
¨ DK/REFUSE
[SKIP TO Q2e]
2b. What type or types of accounts do you and each of your household members have? (Ask this question for each adult
(15 years of age and older) individual of the household.)
¨ Only checking accounts
[CONTINUE]
¨ Only savings accounts
[CONTINUE]
¨ Or both checking and savings accounts
[CONTINUE]
¨ OTHER (VOLUNTEERED)
[CONTINUE]
¨ DK/REFUSE
[CONTINUE]
2e. In the past 12 months, that is since June 2016, was there any time when no one in your household had an account?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
77
2g. In the past 12 months, have you (if OTHERS AGE≥15 FILL: or anyone else in your household) accessed an account in any of
the following ways? (Mark all that apply.)
¨ Bank teller
[CONTINUE]
¨ ATM or bank kiosk
[CONTINUE]
¨ Telephone banking through phone call or automated voice or touch tone
[CONTINUE]
¨ Online banking with a laptop, desktop computer, or tablet such as an iPad
[CONTINUE]
¨ Mobile banking with text messaging, mobile app, or Internet browser or email on a
mobile phone
[CONTINUE]
¨ Other (Specify)
[CONTINUE]
¨ Did not access an account in the past 12 months
[CONTINUE]
¨ DK/REFUSE [CONTINUE]
2h. What was the most common way that you (if OTHERS AGE≥15 FILL: or anyone else in your household) accessed an account?
(Read only answers marked in Q2g. Mark only one.)
¨ Bank teller
[SKIP TO Q70]
¨ ATM or bank kiosk
[SKIP TO Q70]
¨ Telephone banking through phone call or automated voice or touch tone
[SKIP TO Q70]
¨ Online banking with a laptop, desktop computer, or tablet such as an iPad
¨ Mobile banking with text messaging, mobile app, or Internet browser or email on a
mobile phone
¨ Other (Specify)
¨ DK/REFUSE
[SKIP TO Q70]
[SKIP TO Q70]
[SKIP TO Q70]
[SKIP TO Q70]
[Questions 3-7 are asked only of households that do not have a bank account.]
3. Have you (if OTHERS AGE≥15 FILL: or anyone else in your household) ever had a checking or savings account?
¨ YES [CONTINUE]
¨ NO
[SKIP TO Q5]
¨ DK/REFUSE
[SKIP TO Q5]
4. Have you (if OTHERS AGE≥15 FILL: or anyone else in your household) had a checking or savings account in the past 12
months, that is since June 2016?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
5. There are different reasons people might not have a checking or savings account. Do any of the following reasons apply to you
(if OTHERS AGE≥15 FILL: or others in your household)? Do you not have an account…
a1. Because bank hours are inconvenient?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
a2. Because bank locations are inconvenient?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
78 | 2017 FDIC National Survey of Unbanked and Underbanked Households
b1. Because bank account fees are too high?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
b2. Because bank account fees are unpredictable?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
c. Because banks do not offer products or services you need?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
d. Because you don’t trust banks?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
e. Because you do not have enough money to keep in an account?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
f. Because avoiding a bank gives more privacy?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
g. Because you cannot open an account due to personal identification, credit, or former bank account problems?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
h. Was there some other reason?
¨ YES (Specify)
[CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
79
6. What is the main reason why no one in your household has an account? (Read only answers marked in Q5a1-Q5h.
Mark only one.)
¨ Bank hours are inconvenient
[CONTINUE]
¨ Bank locations are inconvenient
[CONTINUE]
¨ Bank account fees are too high
[CONTINUE]
¨ Bank account fees are unpredictable
[CONTINUE]
¨ Banks do not offer products or services you need
[CONTINUE]
¨ Don’t trust banks
[CONTINUE]
¨ Do not have enough money to keep in an account
[CONTINUE]
¨ Avoiding a bank gives more privacy
[CONTINUE]
¨ Cannot open an account due to personal identification, credit, or former bank
account problems
[CONTINUE]
¨ Some other reason (Specify)
[CONTINUE]
¨ DK/REFUSE
[CONTINUE]
7. How likely is it that you (if OTHERS AGE≥15 FILL: or anyone in your household) will open a checking or savings account
within the next 12 months?
¨ Very likely
[CONTINUE]
¨ Somewhat likely
[CONTINUE]
¨ Not very likely
[CONTINUE]
¨ Not at all likely
[CONTINUE]
¨ DK/REFUSE [CONTINUE]
[Question 70 is asked only of households that are unbanked or that are banked but did not access an account using a bank teller.]
70. In the past 12 months, have you (if OTHERS AGE≥15 FILL: or anyone else in your household) spoken with a teller or other
employee in person at a bank branch?
¨ YES [CONTINUE]
¨ NO
[SKIP TO Q80]
¨ DK/REFUSE
[SKIP TO Q80]
[Question 71 is asked only of households that spoke with a bank teller (or other employee) in the past 12 months (answered YES to
Q70 or accessed an account using a bank teller).]
71. How many times have you (if OTHERS AGE≥15 FILL: or someone else in your household) spoken with a teller or other
employee in person at a bank branch in the past 12 months?
¨ 1 to 4 times in the past 12 months
[CONTINUE]
¨ 5 to 9 times in the past 12 months
[CONTINUE]
¨ 10 or more times in the past 12 months
[CONTINUE]
¨ DK/REFUSE [CONTINUE]
80 | 2017 FDIC National Survey of Unbanked and Underbanked Households
[Questions 80a-80b are asked only of households that are banked or recently unbanked. Otherwise, skip to Q110.]
The next questions ask about ways you (if OTHERS AGE≥15 FILL: or anyone else in your household) might have used a
mobile phone for banking activities.
80a. In the past 12 months, have you (if OTHERS AGE≥15 FILL: or anyone else in your household) used a mobile phone to check
email from a bank about an account?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
80b. Received a mobile text alert or push notification from a bank about an account?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
[Questions 80c-80g are asked only of households that used mobile banking to access an account. Otherwise, skip to Q110.]
80c. In the past 12 months, have you (if OTHERS AGE≥15 FILL: or anyone else in your household) used a bank’s mobile website
or bank’s mobile app to check a bank account balance or recent transactions?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
80d. Used a bank’s mobile website or bank’s mobile app to make a bill payment?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
80e. Used a bank’s mobile website or bank’s mobile app to send money to other people?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
80f. Used a bank’s mobile website or bank’s mobile app to transfer money between bank accounts owned by the same person?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
80g. In the past 12 months, have you (if OTHERS AGE≥15 FILL: or anyone else in your household) used a mobile phone’s camera
to deposit a check into a bank account?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
81
Now I have a question about prepaid cards. Prepaid cards allow you or others, like relatives or a government agency,
to load funds that can later be spent. Prepaid cards also allow you to withdraw cash from ATMs. I am not asking about
gift cards or debit cards linked to a checking account.
110. In the past 12 months, that is since June 2016, did you (if OTHERS AGE≥15 FILL: or anyone else in your household) use any
prepaid cards?
¨ YES [CONTINUE]
¨ NO
[SKIP to Q120]
¨ DK/REFUSE
[SKIP to Q120]
[Question 111 is asked only of households that used a prepaid card in the last 12 months.]
111. Where did the prepaid cards that you used in the past 12 months come from? (Mark all that apply.)
¨ A bank location or bank’s website
[CONTINUE]
¨ A store or website that is not a bank
[CONTINUE]
¨ A government agency
[CONTINUE]
¨ Employer payroll card
[CONTINUE]
¨ Family or friends
[CONTINUE]
¨ Other (Specify)
[CONTINUE]
¨ DK/REFUSE [CONTINUE]
[Question 112 is asked only of households that used a prepaid card from a government agency.]
112. Thinking about the card(s) received from a government agency, why did you (if OTHERS AGE≥15 FILL: or others in your
household) have these card(s)? (Mark all that apply.)
¨ To receive Social Security or disability benefits
[CONTINUE]
¨ To receive unemployment benefits
[CONTINUE]
¨ To receive food or child care benefits like SNAP or WIC
[CONTINUE]
¨ Other (Specify)
[CONTINUE]
¨ DK/REFUSE [CONTINUE]
Earlier, we asked about banks, including any bank, savings and loans institution, credit union, or brokerage firm. The
next questions ask about going to places other than a bank for your financial services.
120. In the past 12 months, that is since June 2016, did you (if OTHERS AGE≥15 FILL: or anyone else in your household) go to
some place other than a bank to cash a check?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
121. In the past 12 months, did you (if OTHERS AGE≥15 FILL: or anyone else in your household) go to some place other than a
bank to purchase a money order?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
122. Did you (if OTHERS AGE≥15 FILL: or anyone else in your household) take out a payday loan or payday advance from some
place other than a bank in the past 12 months?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
82 | 2017 FDIC National Survey of Unbanked and Underbanked Households
123. Did you (if OTHERS AGE≥15 FILL: or anyone else in your household) pawn an item at a pawn shop in the past 12 months?
Do not include selling an unwanted item to a pawn shop.
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
124. In the past 12 months, that is since June 2016, did you (if OTHERS AGE≥15 FILL: or anyone else in your household) take
out a tax refund anticipation loan, or use a tax preparation service in order to receive your tax refund faster than the IRS would
provide it?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
126. Auto title loans use a car title to borrow money for a short period of time. They are NOT loans used to purchase a car. In the
past 12 months, did you (if OTHERS AGE≥15 FILL: or someone else in your household) take out an auto title loan?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 127 is asked only of households that indicated NO (or DK/REFUSE) to Q122, Q123, Q124, and Q126.]
127. In the past 12 months, have you (if OTHERS AGE≥15 FILL: or anyone else in your household) taken out any other types of
loans or lines of credit from a payday lender, auto title lender, pawn shop, or check casher?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 125 is asked of all households.]
125. Some stores allow people to rent to own items such as furniture or appliances. We do not mean stores that offer installment
plans or layaway plans. In the past 12 months, did you (if OTHERS AGE≥15 FILL: or anyone else in your household) rent anything
from a rent-to-own store because it couldn’t be financed any other way?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
The next few questions are about sending money abroad.
130. In the last 12 months, that is since June 2016, did you (if OTHERS AGE≥15 FILL: or someone else in your household) send
money to family or friends living outside of the US?
¨ YES
¨ NO
¨ DK/REFUSE
[CONTINUE]
[SKIP TO Q140a]
[SKIP TO Q140a]
[Question 135 is asked only of households that sent money abroad.]
135. When sending money abroad in the last 12 months, did you (if OTHERS AGE≥15 FILL: or someone else in your household)
use… (Mark all that apply.)
¨ A bank?
[CONTINUE]
¨ A place other than a bank?
[CONTINUE]
¨ Other (Specify)
[CONTINUE]
¨ DK/REFUSE [CONTINUE]
83
The next few questions are about the different ways people receive income. People may receive income from work,
retirement, government benefits, or other sources in a number of ways. Think about the ways your household has
received income during a typical month, in the past 12 months.
[Question 140a is asked of all households.]
140a. In a typical month, have you (if OTHERS AGE≥15 FILL: or others in your household) received income by paper check or
money order?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DID NOT RECEIVE INCOME IN PAST 12 MONTHS (VOLUNTEERED)
[SKIP TO Q150a]
¨ DK/REFUSE
[CONTINUE]
[Question 140b is asked only of households that are banked or recently unbanked.]
140b. How about through direct deposit or electronic transfer into a bank account?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 140c is asked only of households that used a prepaid card in the past 12 months.]
140c. In a typical month, have you (if OTHERS AGE≥15 FILL: or others in your household) received income through direct deposit
or electronic transfer onto a prepaid card?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Questions 140d-140e are asked of all households.]
140d. In a typical month, have you (if OTHERS AGE≥15 FILL: or others in your household) received income in cash?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
140e. In a typical month, have you (if OTHERS AGE≥15 FILL: or others in your household) received income in any other form?
¨ YES (Specify)
[CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 140x is asked only of households that indicated NO to all applicable questions in Q140a-Q140e.]
140x. In the past 12 months, did you (if OTHERS AGE≥15 FILL: or others in your household) receive any income from work, retirement, government benefits, or other sources?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
[Question 141 is asked only of households that received income by paper check or money order, and used a nonbank check casher in the last 12 months.]
141. Think about the income you (if OTHERS AGE≥15 FILL: or others in your household) received by paper check or money order
in the past 12 months. Did you typically use some place other than a bank to cash the check or money order?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
84 | 2017 FDIC National Survey of Unbanked and Underbanked Households
The next few questions are about the different ways people pay their monthly bills for things like mortgage, rent, utilities, or child care. Think about the ways your household has paid bills during a typical month, in the past 12 months.
[Question 150a is asked of all households.]
150a. In a typical month, did you (if OTHERS AGE≥15 FILL: or someone else in your household) use cash to pay these types of
bills?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DID NOT PAY BILLS IN PAST 12 MONTHS (VOLUNTEERED)
[SKIP TO Q1600a]
¨ DK/REFUSE
[CONTINUE]
[Questions 150b-150c are asked only of households that are banked or recently unbanked.]
150b. How about using a personal check drawn on a bank account to pay bills?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
150c. How about using a debit card linked to a bank account to pay bills?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 150d is asked of all households.]
150d. In a typical month, did you (if OTHERS AGE≥15 FILL: or someone else in your household) use a credit card to pay bills?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 150e is asked only of households that used a prepaid card in the last 12 months.]
150e. In a typical month, did you (if OTHERS AGE≥15 FILL: or someone else in your household) use a prepaid card to pay bills?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 150f is asked only of households that are banked or recently unbanked.]
150f. In a typical month, did you (if OTHERS AGE≥15 FILL: or someone else in your household) pay bills electronically from a bank
account, either through online bill pay or direct withdrawal?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 150g is asked only of households that used a money order from a place other than a bank in the last 12 months.]
150g. In a typical month, did you (if OTHERS AGE≥15 FILL: or someone else in your household) use a money order from a place
other than a bank to pay bills?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
85
[Questions 150h-150i are asked of all households.]
150h. How about using a cashier’s check or money order from a bank to pay bills?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
150i. In a typical month, did you (if OTHERS AGE≥15 FILL: or someone else in your household) pay bills in any other way?
¨ YES (Specify)
[CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 150x is asked only of households that indicated NO to all applicable questions in Q150a-Q150i.]
150x. In the past 12 months, did you (if OTHERS AGE≥15 FILL: or others in your household) pay any bills for things like mortgage,
rent, utilities, or child care?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
The next few questions are about how people borrow money or purchase items on credit.
[Questions 1600a-1600g are asked of all households.]
1600. In the past 12 months, have you (if OTHERS AGE≥15 FILL: or anyone else in your household) had any of the following?
a. A credit card from Visa, MasterCard, American Express, or Discover? Please do not include debit cards.
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
b. A store credit card that can only be used at that store? Please do not include gift cards.
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
c. In the past 12 months, have you (if OTHERS AGE≥15 FILL: or anyone else in your household) had an auto loan? [If YES to
Q126, then FILL: This is different from an auto title loan.]
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
d. A mortgage or home equity loan or home equity line of credit?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
e. In the past 12 months, have you (if OTHERS AGE≥15 FILL: or anyone else in your household) had a student loan?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
86 | 2017 FDIC National Survey of Unbanked and Underbanked Households
f. Other personal loans or lines of credit from a bank?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
g. Other personal loans or lines of credit from a company other than a bank? [If YES to Q124, then FILL: Please do not include
refund anticipation loans or any loans from a payday lender, pawn shop, auto title lender, or check casher.] [If Q124 is NOT YES
and any of Q122, Q123, Q126, or Q127 are YES, then FILL: Please do not include any loans from a payday lender, pawn shop,
auto title lender, or check casher.]
¨ YES (Specify who provided the loan)
[CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
[Question 162 is asked of all households.]
162. In the past 12 months, that is since June 2016, did you (if OTHERS AGE≥15 FILL: or someone else in your household) apply
for a new credit card, or a personal loan or line of credit at a bank?
¨ YES [CONTINUE]
¨ NO
[SKIP TO Q164]
¨ DK/REFUSE
[SKIP TO Q164]
[Question 163 is asked only of households that applied for credit in the last 12 months.]
163. In the past 12 months, did any lender or creditor turn down your (if OTHERS AGE≥15 FILL: or someone else in your household’s) request for new credit or not give you as much credit as you applied for?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
[Question 164 is asked of all households.]
164. Was there any time in the past 12 months that you (if OTHERS AGE≥15 FILL: or someone else in your household) thought
about applying for a new credit card, or a personal loan or line of credit at a bank, but changed your mind because you thought
you might be turned down?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE
[CONTINUE]
The next few questions are about the different ways that people save their money.
170. Even if you later spent it, did you (if OTHERS AGE≥15 FILL: or anyone in your household) set aside any money in the past 12
months that could be used for unexpected expenses or emergencies? I’m only asking about funds that could be easily spent if
necessary, and am not asking about retirement or other long-term savings.
¨ YES [CONTINUE]
¨ NO
[SKIP TO Q180]
¨ DK/REFUSE
[SKIP TO Q180]
[Question 171 is asked only of households that set aside some savings in the past 12 months.]
87
171. Where did you (if OTHERS AGE≥15 FILL: or anyone in your household) keep this money? (Mark all that apply.)
¨ (Read only for banked or recently unbanked) In a checking account?
[CONTINUE]
¨ (Read only for banked or recently unbanked) In a savings account?
[CONTINUE]
¨ (Read only for those with a prepaid card) On a prepaid card?
[CONTINUE]
¨ In other accounts such as certificates of deposit, brokerage accounts,
or savings bonds?
[CONTINUE]
¨ Did you keep the savings in the home, or with family or friends?
[CONTINUE]
¨ Did you buy something with the intent to pawn or sell later if necessary?
[CONTINUE]
¨ Other (Specify)
[CONTINUE]
¨ DK/REFUSE [CONTINUE]
[Questions 180-181 and 185 are asked of all households.]
180. Which best describes your household’s income over the past 12 months? (Mark only one.)
¨ Income is about the same each month
¨ Income varies somewhat from month to month
¨ Income varies a lot from month to month
¨ DK/REFUSE
[CONTINUE]
[CONTINUE]
[CONTINUE]
[CONTINUE]
181. Often times, households find that they are not able to keep up with their bills. Over the last 12 months, was there a time
when you (if OTHERS AGE≥15 FILL: or someone else in your household) fell behind on bill payments?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
185. Do you (if OTHERS AGE≥15 FILL: or someone else in your household) currently own or have regular access to a mobile
phone?
¨ YES [CONTINUE]
¨ NO
[SKIP TO Q187]
¨ DK/REFUSE
[SKIP TO Q187]
[Question 186 is asked only of households that have a mobile phone.]
186. Are any of these mobile phones a smartphone with features to access the Internet, send emails, and download apps?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
[Question 187 is asked of all households.]
187. Do you (if OTHERS AGE≥15 FILL: or someone else in your household) currently have regular access to the Internet
at home using a desktop, laptop, or tablet computer?
¨ YES [CONTINUE]
¨ NO [CONTINUE]
¨ DK/REFUSE [CONTINUE]
88 | 2017 FDIC National Survey of Unbanked and Underbanked Households
Federal Deposit Insurance Corporation
FDIC-038-2018
| File Type | application/pdf |
| File Modified | 0000-00-00 |
| File Created | 0000-00-00 |