This is the accessible text file for GAO report number GAO-09-8
entitled 'Check 21 Act: Most Consumers Have Accepted and Banks Are
Progressing Toward Full Adoption of Check Truncation' which was
released on October 28, 2008.
This text file was formatted by the U.S. Government Accountability
Office (GAO) to be accessible to users with visual impairments, as part
of a longer term project to improve GAO products' accessibility. Every
attempt has been made to maintain the structural and data integrity of
the original printed product. Accessibility features, such as text
descriptions of tables, consecutively numbered footnotes placed at the
end of the file, and the text of agency comment letters, are provided
but may not exactly duplicate the presentation or format of the printed
version. The portable document format (PDF) file is an exact electronic
replica of the printed version. We welcome your feedback. Please E-mail
your comments regarding the contents or accessibility features of this
document to Webmaster@gao.gov.
This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed
in its entirety without further permission from GAO. Because this work
may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this
material separately.
Report to Congressional Committees:
United States Government Accountability Office:
GAO:
October 2008:
Check 21 Act:
Most Consumers Have Accepted and Banks Are Progressing Toward Full
Adoption of Check Truncation:
GAO-09-8:
GAO Highlights:
Highlights of GAO-09-8, a report to Congressional Committees.
Why GAO Did This Study:
Although check volume has declined, checks still represent a
significant volume of payments that need to be processed, cleared, and
settled. The Check Clearing for the 21st Century Act of 2003 (Check 21)
was intended to make check collection more efficient and less costly by
facilitating wider use of electronic check processing. It authorized a
new legal instrument—the substitute check—a paper copy of an image of
the front and back of the original check. Check 21 facilitated
electronic check processing by allowing banks to use electronic imaging
technology for collection and create substitute checks from those
images for delivery to banks that do not accept checks electronically.
Check 21 mandated that GAO evaluate the implementation and
administration of the act. The report objectives are to (1) determine
the gains in economic efficiency from check truncation and evaluate the
benefits and costs to the Federal Reserve System (Federal Reserve) and
financial institutions; (2) assess consumer acceptance of the check
truncation process resulting from Check 21; and (3) evaluate the
benefits and costs to bank consumers from check truncation. GAO
analyzed costs for the check operations of the Federal Reserve and a
group of banks, interviewed consumers about their acceptance of and
costs and benefits of electronic check processing, and analyzed survey
data on bank fees.
The Federal Reserve agreed with the overall findings of the report.
What GAO Found:
Check truncation has not yet resulted in overall gains in economic
efficiency for the Federal Reserve or for a sample of banks while
Federal Reserve and bank officials expect efficiencies in the future.
GAO’s analysis of the Federal Reserve’s cost accounting data suggests
that its costs for check clearing may have increased since Check 21,
which may reflect that the Federal Reserve must still process paper
checks while it invests in equipment and software for electronic
processing and incurs costs associated with closing a number of check
offices. However, GAO found that the Federal Reserve’s work hours and
transportation costs associated with check services declined from the
fourth quarter of 2001 through the fourth quarter of 2007. Several of
the 10 largest U.S. banks reported to GAO that maintenance of both
paper and image-based check processing systems prevented them from
achieving overall lower costs, although they had reduced transportation
and labor costs since Check 21 was enacted. Check imaging and the use
of substitute checks appear to have had a neutral or minimal effect on
bank fraud losses.
Most bank consumers seem to have accepted changes to their checking
accounts from check truncation. In interviews with bank consumers, the
majority of them accepted not receiving their canceled checks and being
able to access information about their checking account activity
online. Several reported that they did not need the “extra paper” from
canceled checks and that image statements and online reviewing was more
secure than receiving canceled checks. Eleven percent of the 108
consumers still preferred to receive canceled checks. Most consumers
reported that they were not significantly concerned about their ability
to demonstrate proof of payment using a substitute check or check image
rather than a canceled check and few reported that they suffered errors
from the check truncation process. Also, GAO found that the federal
banking regulators reported few consumer complaints relating to Check
21.
To the extent that banks have employed check truncation, bank consumers
have realized benefits and costs relating to faster processing and
access to account information. GAO found that some banks have extended
the hours for accepting deposits for credit on the same business day,
which can result in faster availability of deposited funds for
consumers. Based on consumer interviews, consumers have benefited from
receiving simpler imaged account statements and immediate access to
information about check payments. Check 21’s expedited recredit (prompt
investigation of claims that substitute checks were improperly charged
to accounts and recrediting of the amount in question) also is
considered a consumer benefit. However, based on our consumer and bank
interviews, it appears that a small number of consumers have filed
expedited recredit claims. Based on analysis of survey data on bank
fees, GAO found some consumers may incur fees related to receiving
canceled checks and images. Since 2004, fees for canceled checks appear
to have increased, while fees for images appear to have remained
relatively flat.
What GAO Recommends:
To view the full product, including the scope and methodology, click on
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-09-8]. To view the E-
supplement, click on GAO-09-9SP. For more information, contact Yvonne
D. Jones at (202) 512-8678 or jonesy@gao.gov.
[End of section]
Contents:
Letter:
Results in Brief:
Background:
Check Truncation Has Not Resulted Yet in Overall Gains in Economic
Efficiency for the Federal Reserve or for Banks, but Is Expected to
Produce Efficiencies in the Future:
Most Bank Consumers Appeared to Have Accepted Changes to Their Checking
Accounts from the Check Truncation Process Resulting from Check 21:
Bank Consumers Have Realized Benefits and Costs Relating to Faster
Check Processing and Access to Information about Their Checking
Accounts:
Agency Comments and Our Evaluation:
Appendix I: Objectives, Scope, and Methodology:
Appendix II: Econometric Analysis of Check 21 for Economic Efficiency
in the Federal Reserve's Check Services:
Appendix III: Comments from the Board of Governors of the Federal
Reserve System:
Appendix IV: GAO Contact and Staff Acknowledgments:
Tables:
Table 1: Demographic Characteristics of Consumers Interviewed:
Table 2: Definition of Institution Size Categories:
Table 3: Summary Statistics of Some Selected Variables:
Table 4: Estimation of Logarithm of Total Check Operating Cost, 1994-
2007:
Figures:
Figure 1: Paper-Based Check Collection and Processing:
Figure 2: Example of a Substitute Check:
Figure 3: Check Image Processing:
Figure 4: Distribution of the Number of Noncash Payments:
Figure 5: Number of Check Images Deposited and Received and Number of
Substitute Checks Printed, June 2006-June 2008:
Figure 6: Federal Reserve's Transportation Costs for Check Services,
1994-2007:
Figure 7: Federal Reserve Total Work Hours in Check Services, 1994-
2007:
Figure 8: Bank Consumer Preferences for Reviewing Check Payments
Activity:
Figure 9: Bank Consumer Concern about Demonstrating Proof of Payment
Using a Substitute Check or Image Statement:
Figure 10: Errors Reported by Bank Consumers Involving Canceled Checks
and Image Statements:
Figure 11: Check Enclosure and Imaging Fees, Banks and Savings and
Loans, 2001-2006:
Abbreviations:
ABA: American Bankers Association:
ACH: automated clearing house:
ATM: automated teller machine:
BEA: Bureau of Economic Analysis:
BLS: Bureau of Labor Statistics:
Check 21: Check Clearing for the 21ST Century Act of 2003:
EBT: electronic benefits transfer:
ECCHO: Electronic Check Clearing House Organization:
EFAA: Expedited Funds Availability Act of 1987:
Federal Reserve: Federal Reserve System:
Federal Reserve Board: Board of Governors of the Federal Reserve
System:
ICL: image cash letter:
Informa: Informa Research Services:
United States Government Accountability Office:
Washington, DC 20548:
October 28, 2008:
Congressional Committees:
In the last 10 years technological innovations and consumer and
business preferences for electronic payments have transformed the U.S.
retail payments system from a largely paper-based system to one that
mostly uses electronic transactions. In 2007, the Federal Reserve
System (Federal Reserve) reported that electronic payments, including
credit and debit cards, exceeded two-thirds of all noncash payments
while the number of checks written declined from more than 37 billion
checks in 2003 to 33 billion checks in 2006.[Footnote 1] Although check
volume has declined, paper checks still represent a significant number
of payments that need to be processed, cleared, and settled. The paper-
based collection system for checks has been a labor-intensive process
because at each step in the collection process, the paper check has had
to be physically handled and transported before being settled.
The Check Clearing for the 21st Century Act (Check 21), enacted in
2003, was intended to make the check payment system more efficient and
less costly by facilitating wider use of electronic check processing
without demanding that any bank change its current check collection
practices.[Footnote 2] At the time that Check 21 was enacted, most
banks could not leverage their investments in imaging technology to
collect checks electronically because the legal framework for the check
collection system constrained the efforts of many banks to use
it.[Footnote 3] Prior to Check 21, a bank was required to present an
original paper check to the bank where the check was payable--the
paying bank--for payment unless the paying bank had agreed to accept
presentment in some other form. This required the bank presenting the
check--the collecting bank--to enter into agreements with all or nearly
all of the banks to which it presented checks. Because of this
impediment, banks were deterred from making the necessary investments
to collect checks electronically.
Check 21 addressed this situation by authorizing a new paper negotiable
instrument, called a substitute check, that when properly prepared is
the legal equivalent of the original check. Any bank that transfers,
presents, or returns a substitute check warrants that the substitute
check contains an accurate image of the front and the back of the
original check at the time the original check was truncated and a
specific legend stating that the substitute check is a legal copy of
the original check and can be used in the same way one would use the
original check. Check 21 does not require the banks to adopt electronic
check processing, but enables banks that want to truncate or remove the
original paper checks from the check-collection system to do so. Check
21 facilitated electronic processing by allowing banks to use
electronic imaging technology for collection and create substitute
checks from those images for delivery to banks that do not accept
checks electronically. Substitute checks are considered an intermediate
step toward a matured electronic check processing system, in which the
goal should be the electronic exchange of payment information and check
images between banks.
Check 21 mandated that we evaluate the implementation and
administration of Check 21. To respond to the Check 21 mandate, the
objectives of this report are to (1) determine the gains in economic
efficiency from check truncation and evaluate the benefits and costs to
banks and the Federal Reserve from check truncation;[Footnote 4] (2)
assess consumer acceptance of the check truncation process resulting
from Check 21; and (3) evaluate the benefits and costs to consumers
from check truncation under Check 21.
To determine the gains in economic efficiency from check truncation and
evaluate the benefits and costs to financial institutions from check
truncation, we separately analyzed costs for the check operations of
the Federal Reserve and a group of banks. Using data from the Federal
Reserve's cost accounting system, we applied an econometric cost model
to estimate the effects of different variables, such as the volume of
checks processed, wages, and other costs incurred by the Federal
Reserve, on total check processing costs from 1994 through 2007. While
the Federal Reserve has consistent cost accounting data, the banking
industry does not. Accounting for costs associated with check
processing varies across the banking industry, preventing a similar
analysis for private-sector costs. Instead, we sent a data collection
instrument to and interviewed officials from the 10 largest banks by
deposit size as of March 2008 in the United States and a group of
smaller banks. The 10 banks account for a significant volume of checks
presented (in 2007, about one-third of all checks paid). We asked about
costs related to paper check processing, the investment that banks
incurred to exchange check images, the cost savings that banks achieved
(including labor and transportation) with image technology, and the
impact of check imaging and substitute checks on losses from fraudulent
checks. We sent the data collection instrument to the 10 banks and
received a response from 9. For the bank that did not respond, we
interviewed an official representing the bank at an early stage of our
engagement. We conducted follow-up interviews with a number of the
institutions requesting clarification of their responses. In addition,
we sent the data collection instrument to 12 smaller banks, which had
assets ranging from less than $500 million to $5 billion. Our selection
criteria included whether the banks were located in metropolitan or
nonmetropolitan areas and were on the Electronic Check Clearing House
Organization's (ECCHO) list of participating members. From this group
of 12 banks, we received five completed forms. We conducted follow-up
interviews with three of the smaller banks. To assess bank consumer
acceptance of the check truncation process resulting from Check 21, we
conducted in-person interviews with 108 consumers. The consumers
represented an approximate distribution of the U.S. adult population
across broad categories (age, education, and income). Consumers had to
meet certain other conditions: having primary responsibility in the
household for balancing the financial account that allows paper check
writing and having received canceled original checks in paper form with
the checking account statement at some point since 2000. However, the
consumers recruited for the interviews did not form a random,
statistically representative sample of the U.S. population; therefore,
we could not generalize the results of the interviews to the relevant
total population. The interview questions covered topics, such as how
consumers reviewed their checking account activity, their acceptance of
the check truncation process, and any problems or errors they might
have had with their checking accounts since Check 21. This report does
not contain all the results from the consumer interviews. We reproduced
the text from our structured interview instrument and tabulated the
results from the questions in Questions for Consumers about Check 21
Act (GAO-09-9SP). To evaluate the benefits and costs to bank consumers
from check truncation, we interviewed Federal Reserve Board staff,
representatives from Consumers Union, the Consumer Federation of
America, and the U.S. Public Interest Research Group, and bank
officials to identify possible benefits and costs. We also analyzed a
study by the Board of Governors of the Federal Reserve System (Federal
Reserve Board) that assessed the banking industry's implementation of
Check 21. To determine whether bank consumers incurred fees for
receiving canceled checks and check images since Check 21, we reviewed
survey data on bank fees for 2001 through 2006 collected by Informa
Research Services Inc., a private-sector firm.
We conducted this performance audit from September 2007 to October 2008
in accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objectives. We believe that
the evidence obtained provides a reasonable basis for our findings and
conclusions based on our audit objectives. See appendix I for a more
detailed discussion of our scope and methodology.
Results in Brief:
Check truncation has not yet resulted in overall gains in economic
efficiency for the Federal Reserve or for selected banks, but Federal
Reserve and bank officials expect efficiencies in the future. The
expectation for electronic processing of checks was that it would lead
to gains in economic efficiency--that is, removing paper from the
payment stream would lead to lower costs. However, our analysis of
Federal Reserve cost accounting data does not demonstrate that the
Federal Reserve's costs for processing payments have decreased--which
may reflect that the Federal Reserve must still maintain its ability to
process paper checks, while it invests in equipment and software for
electronic check processing and incurs costs associated with closing a
number of check processing sites. But, we also found that the total
work hours associated with the Federal Reserve's check processing
operations decreased by approximately 48 percent from 2.1 million hours
for the fourth quarter of 2001 to 1.3 million hours for the fourth
quarter of 2007 and the transportation costs associated with check
processing operations decreased by about 11 percent from the fourth
quarter of 2001 to the fourth quarter of 2007. Estimates of whether
costs were lower for banks as a result of Check 21 varied considerably,
reflecting the diverse ways in which they handle checks and payments
and differences among cost accounting systems. For example, several of
the 10 largest banks noted that maintaining a dual paper-electronic
infrastructure to date had prevented them from achieving overall lower
costs, although they also had seen reduced transportation and labor
costs. As Federal Reserve officials noted, the willingness of private
banks to invest in the equipment needed to process checks
electronically demonstrated the bank's expectation of lower costs. The
banks said that they expect that eventually costs would be lower.
According to our interviews with three smaller banks, they generally
have migrated all of their check volumes to electronic processing
rather than operating two processing systems and have seen lower costs
for transportation and labor. Since cost accounting systems vary among
banks and many were unwilling to share proprietary data on their costs,
it was not possible to estimate the industrywide cost effect of check
truncation. Check imaging and the use of substitute checks appear to
have had a neutral or minimal effect on bank fraud losses.
Most bank consumers seem to have accepted changes to their checking
accounts from the check truncation process. In one-on-one, in-person
structured interviews with 108 bank consumers living in three cities,
we learned that the majority of these consumers accepted not receiving
their canceled checks and being able to access information about their
checking account activity online. Eleven percent of these consumers
still wanted to receive canceled checks with their account statement.
In addition, most consumers reported that they were not concerned
significantly about their ability to demonstrate proof of payment using
a substitute check or check image rather than a canceled check. Few
consumers reported that they experienced errors from the check
truncation process. In addition, we found on the basis of our review of
consumer complaints that the federal banking regulators reported few
complaints about Check 21. Of the approximately 35,000 consumer
complaints submitted to the four federal banking regulators in 2006 and
2007, 172 were related to Check 21 issues. The primary consumer
complaint was that the account holder wanted to continue receiving
canceled checks. These findings appear consistent with findings in the
Federal Reserve Board's April 2007 report to Congress on Check 21.
Specifically, the Federal Reserve Board reported that less than 1
percent of all complaints received by the federal banking regulators
were related to Check 21.
To the extent that banks have implemented electronic check processing,
bank consumers have realized both benefits and costs relating to faster
processing and access to information about their checking accounts. We
found that some banks reported that they have extended the cut-off time
for accepting deposits for credit on the same business day, which can
result in faster availability of deposited funds for consumers.
However, the funds availability requirements of the Federal Reserve
Board's Regulation CC have not been amended as a result of Check 21.
Regulation CC limits the time that banks can hold funds deposited into
customer accounts before these funds must be made available for
withdrawal. In its April 2007 report, the Federal Reserve Board
concluded that much broader adoption of new technologies and processes
by the banking industry must occur before check return times could
decline appreciably and thereby permit a modification of the funds
availability deadlines. However, Check 21 has contributed to
acceleration of the pace of the Federal Reserve's consolidation of
check processing offices, which has increased the proportion of checks
that are classified as local. As a result, a large number of consumer
checks are now subject to shorter maximum hold periods by banks under
Regulation CC. Many bank consumers have realized other benefits
including simpler statements (that is, consumers said it was easier to
review images on a few sheets of paper than handle many canceled
checks) and immediate access to information about payments. Bank and
industry association officials also noted benefits of check truncation,
such as better customer access to check images and accelerated customer
deposit availability. Check 21's expedited recredit provision
(requiring banks to complete their investigation of a consumer's claim
that a substitute check had been improperly charged to their account
within 10 business days and recredit the consumer's account for a
specified amount pending completion of the investigation) is considered
a consumer benefit. However, on the basis of our consumer and bank
interviews, it appears that a small number of consumers have filed
expedited recredit claims. Some consumers also may incur fees if they
elect to receive canceled checks and check images. Based on a survey of
large retail banks conducted by Informa Research Services from 2001 to
2006, fee amounts for receiving canceled checks generally have
increased, while fees for image statements appear to have remained
relatively flat. The Informa data also indicated that these banks
charge different amounts depending on the type of checking account.
We provided a copy of a draft of this report to the Federal Reserve
Board, which provided us with written comments that are reprinted in
appendix III. It agreed with our overall conclusion that, over the past
four years, the banking industry has made substantial progress toward
establishing an end-to-end electronic check-processing environment. In
commenting on this report, it noted that the Federal Reserve Banks
expect that by year-end 2009, more than 90 percent of their check
deposits and presentments will be electronic. They also commented that
the ongoing transformation to electronic check-processing environment
has not been without cost. As noted in our report, the Federal Reserve
Banks have reduced their transportation costs and work hours associated
with their check services. And, according to the Federal Reserve Board,
they earned a net income of $326 million for providing check services
from 2005 through 2007. The Federal Reserve Board concurred with a
number of consumer benefits identified in the report: faster funds
availability on check deposits due to later deposit deadlines, quicker
access to account information, and improved customer service. In
addition, we sent a draft of this report to the Federal Deposit
Insurance Corporation, Office of the Comptroller of the Currency, and
Office of Thrift Supervision. Only the Office of the Comptroller of the
Currency provided us with technical comments, which we incorporated as
appropriate. We provided sections of the draft of this report to bank
officials for their technical review and several of them provided us
technical comments, which we incorporated as appropriate.
Background:
This section of the report describes the paper-and electronic-based
check collection processes, presents statistics on the use of
electronic and nonelectronic payments and types of check processing,
and describes the Federal Reserve's role in check collection.
Check Collection Process:
Interbank checks are cleared and settled through an elaborate check-
collection process that includes presentment and final
settlement.[Footnote 5] Check presentment occurs when the checks are
delivered or images transmitted to the paying banks for payment and the
paying banks must decide whether to honor or return the checks (see
fig. 1). Settlement of checks occurs when the collecting banks are
credited and the paying banks are debited, usually through accounts
held at either the Federal Reserve or correspondent banks.
Figure 1: Paper-Based Check Collection and Processing:
This figure is a combination of illustrations showing paper-based check
collection and processing.
[See PDF for image]
Source: GAO (analysts); Art Explosion (images).
[End of figure]
In the paper-based check collection process, banks of first deposit
generally sort deposited checks by destination and dispatch them for
collection. Banks of first deposit physically can collect a paper check
through several methods:
* Direct presentment of the paper check to the paying bank;
* Exchange of the paper check at a clearing house in which the bank of
first deposit and the paying bank are members;
* Collection of the paper check through an intermediary, such as a
correspondent bank or a Federal Reserve Bank; or:
* Some combination of the above methods.
When a paying bank decides not to pay a check, the bank typically
returns the dishonored check to the bank of first deposit. Under the
Uniform Commercial Code, the paying bank generally has until midnight
of the day following presentment ("midnight deadline") to return
dishonored checks or send notices of dishonor.[Footnote 6] The paying
bank may return a dishonored check, commonly referred to as a return
item, directly to the bank of first deposit through a clearing house
association, if applicable, or through a returning bank (a bank
handling a returned check), including the Federal Reserve.
Regulation CC was promulgated by the Federal Reserve Board in 1988 to
implement the Expedited Funds Availability Act of 1987 (EFAA), which
establishes the maximum periods of time that banks can hold funds
deposited into accounts before those funds must be made available for
withdrawal. Among other things, the EFAA and its implementing
Regulation CC generally require banks to make funds from local checks
available by the second business day after the day of deposit; funds
from nonlocal checks must be available by the fifth business day after
the day of deposit.[Footnote 7]
At each step, the check must be processed physically and then shipped
to its destination by air or ground transportation. Some have suggested
that truncating paper checks, or stopping them before they reach the
paying bank, could result in lower costs to process checks and benefits
to both the banking industry and the public. Under Regulation CC, the
term "truncate" means to remove an original check from the collection
or return process. Instead, the recipient receives a substitute check;
or by agreement, information relating to the original check (including
data taken from the magnetic ink character recognition line of the
original check or an electronic image of the original check), whether
with or without the subsequent delivery of the original check (see fig.
2).[Footnote 8]
Figure 2: Example of a Substitute Check:
This figure is a picture of a substitute check.
[See PDF for image]
Source: ECCHO.
[End of figure]
Electronic Check Processing and Imaging Technology:
Essentially, check imaging is a process through which a paper check is
scanned and a digital image is taken of the front and back of the paper
check. The paper check may then at some point be destroyed and the
images may then be stored in an archive maintained by the bank for
retrieval if needed. When a paper check is imaged depends on the
structure of a bank's back office operations. Some banks have the
capability to image a paper check at their branches, while others
transport the paper to centralized locations where the paper is imaged.
Once the images are taken, an image cash letter (ICL) is assembled and
sent to the paying bank directly or to an intermediary (such as the
Federal Reserve, a correspondent bank, or an image exchange processor)
for ultimate presentment to the paying bank (see fig. 3). Since Check
21 was enacted, imaging technology has been further refined so that it
is possible for a bank to image a paper check at its branches or
automated teller machines (ATM)--commonly referred to as branch or ATM
capture. In addition, some banks are beginning to offer a service to
their customers called remote deposit capture where merchants can scan
the paper checks they receive and electronically deposit those images
at the bank.
Figure 3: Check Image Processing:
This figure is a combination of illustrations of check image
processing.
[See PDF for image]
Source: GAO (analysts); Art Explosion (images).
[End of figure]
As discussed in the introduction to this report, electronic check
processing was hampered by certain legal impediments that Check 21
addressed. Moreover, as we reported in 1998, perceptions about consumer
preferences for receiving canceled checks also deterred electronic
check processing.[Footnote 9] Because, under Check 21, checks drawn on
any particular bank can be truncated by any bank across the country,
banks cannot return the original canceled paper checks to their
customers once they are imaged. At the time of our 1998 report, Federal
Reserve officials and bank officials with whom we spoke expressed a
belief that many consumers wanted their canceled checks
returned.[Footnote 10]
Recent Trends in Check Use, Overall Electronic Payments, and Electronic
Processing of Checks:
The popularity of the paper check as a retail payment instrument in the
United States is waning. The Federal Reserve has estimated that the
number of checks used in the United States peaked during the mid-1990s
at around 50 billion checks per year.[Footnote 11] In its 2007 study
the Federal Reserve highlighted the decline in check usage as a retail
payment instrument. It reported that both the number of checks written
and checks paid declined from 2003 through 2006. In 2006, 33.1 billion
checks were written compared with 37.6 billion checks in 2003 and paid
checks decreased from 37.3 billion checks to 30.6 billion checks in the
same period. The number of checks written differs from checks paid
because paper checks that have been converted into automated clearing
house (ACH) payments were included in the figure for checks
written.[Footnote 12] Additionally, the Federal Reserve concluded that
the share of retail payments made electronically was growing, while the
share of check payments of total noncash payments was declining.
Electronic payments, including debit and credit cards, ACH payments
(including check conversions), and electronic benefit transfers (EBT)
amounted to two-thirds of the total number of noncash payments, which
in 2006 totaled 93.3 billion.[Footnote 13] The share of check payments
declined from 46 percent in 2003 to 33 percent in 2006 (see fig. 4).
Figure 4: Distribution of the Number of Noncash Payments:
This figure is a combination of two pie graphs showing distribution of
the number of noncash payments.
2003:
Checks (paid): 46%;
Credit card: 23%;
Debit card: 19%;
ACH: 11%;
EBT: 1%.
2006:
Checks (paid): 33%;
Credit card: 23%;
Debit card: 27%;
ACH: 16%;
EBT: 1%.
[See PDF for image]
Source: The 2007 Federal Reserve System Payment Study.
[End of figure]
While check use has declined, check processing increasingly has become
electronic. As shown in figure 5, from June 2006 through June 2008, the
number of imaged checks deposited by collecting banks and received by
paying banks has grown steadily. In June 2006 banks deposited 206
million checks as images compared with June 2008, when banks deposited
1.1 billion checks. Similarly, the number of checks received as images
by the paying banks has grown. In June 2006, paying banks received 89
million items; by June 2008, they received almost 852 million items.
However, the number of substitute checks has not declined, but has
increased from 117 million in June 2006 to 283 million in June 2008.
These checks represent paper that must be presented physically to
paying banks through the collection system.
Figure 5: Number of Check Images Deposited and Received and Number of
Substitute Checks Printed, June 2006-June 2008:
This figure is a combination line graph showing the number of check
images deposited and received and number of substitute checks printed,
June 2006-June 2008. The X axis represents month and year, and the Y
axis represents the number of millions. The lines represent checks sent
for collection via image, image items received by paying institution,
and substitute checks received by paying institution.
2006: 06;
Image items received by paying institution: 89;
Substitute checks received by paying institution: 117.4;
Checks sent for collection via image: 206.4.
2006: 07;
Image items received by paying institution: 95.9;
Substitute checks received by paying institution: 120.9;
Checks sent for collection via image: 216.8.
2006: 08;
Image items received by paying institution: 129.7;
Substitute checks received by paying institution: 143.7;
Checks sent for collection via image: 273.3.
2006: 09;
Image items received by paying institution: 143.7;
Substitute checks received by paying institution: 149.6;
Checks sent for collection via image: 293.2.
2006: 10;
Image items received by paying institution: 204.8;
Substitute checks received by paying institution: 184.2;
Checks sent for collection via image: 389.
2006: 11;
Image items received by paying institution: 245.5;
Substitute checks received by paying institution: 182.1;
Checks sent for collection via image: 427.5.
2006: 12;
Image items received by paying institution: 286.6;
Substitute checks received by paying institution: 193.9;
Checks sent for collection via image: 480.5.
2007: 01;
Image items received by paying institution: 319.4;
Substitute checks received by paying institution: 211.1;
Checks sent for collection via image: 530.5.
2007: 02;
Image items received by paying institution: 311.9;
Substitute checks received by paying institution: 193.3;
Checks sent for collection via image: 505.2.
2007: 03;
Image items received by paying institution: 393.2;
Substitute checks received by paying institution: 228.5;
Checks sent for collection via image: 621.7.
2007: 04;
Image items received by paying institution: 446;
Substitute checks received by paying institution: 231.5;
Checks sent for collection via image: 677.5.
2007: 05;
Image items received by paying institution: 501.3;
Substitute checks received by paying institution: 259.3;
Checks sent for collection via image: 760.6.
2007: 06;
Image items received by paying institution: 487.1;
Substitute checks received by paying institution: 247.2;
Checks sent for collection via image: 734.3.
2007: 07;
Image items received by paying institution: 555.4;
Substitute checks received by paying institution: 261.5;
Checks sent for collection via image: 816.9.
2007: 08;
Image items received by paying institution: 598.5;
Substitute checks received by paying institution: 277.1;
Checks sent for collection via image: 875.5.
2007: 09;
Image items received by paying institution: 576.2;
Substitute checks received by paying institution: 263.2;
Checks sent for collection via image: 839.4.
2007: 10;
Image items received by paying institution: 683.4;
Substitute checks received by paying institution: 310;
Checks sent for collection via image: 993.4.
2007: 11;
Image items received by paying institution: 673.6;
Substitute checks received by paying institution: 278.7;
Checks sent for collection via image: 952.4.
2007: 12;
Image items received by paying institution: 716.8;
Substitute checks received by paying institution: 284.8;
Checks sent for collection via image: 1001.6.
2008: 01;
Image items received by paying institution: 745.4;
Substitute checks received by paying institution: 293.9;
Checks sent for collection via image: 1039.3.
2008: 02;
Image items received by paying institution: 713.7;
Substitute checks received by paying institution: 276.8;
Checks sent for collection via image: 990.4.
2008: 03;
Image items received by paying institution: 795.5;
Substitute checks received by paying institution: 292.6;
Checks sent for collection via image: 1088.1.
2008: 04;
Image items received by paying institution: 854.4;
Substitute checks received by paying institution: 306.5;
Checks sent for collection via image: 1160.9.
2008: 05;
Image items received by paying institution: 821.9;
Substitute checks received by paying institution: 285;
Checks sent for collection via image: 1106.8.
2008: 06;
Image items received by paying institution: 851.5;
Substitute checks received by paying institution: 283.3;
Checks sent for collection via image: 1134.8.
[See PDF for image]
Source: GAO analysis of ECCHO data.
[End of figure]
Federal Reserve's Role in Check Collection and Its Consolidation of
Check Offices:
The Federal Reserve operates a comprehensive, nationwide system for
clearing and settling checks drawn on banks located throughout the
United States. These offices accept paper check deposits and transport
the paper checks to the paying bank. Since the effective date of Check
21, the Federal Reserve sends and receives images between banks. The
Federal Reserve offers imaged check products--commonly referred to as
the Check 21 products (Fed Forward, Fed Receipt, and Fed Return)--for a
fee to banks that use its check collection services.[Footnote 14]
According to the Federal Reserve Board's 2007 Annual Report, of the
approximately 10 billion checks (about one-third of the total 30.6
billion paid checks) processed through the Federal Reserve in 2007,
42.2 percent were deposited as images and 24.6 percent were received
using Check 21 products. Further, in the month of July 2008, the
proportion of checks deposited and presented as images using the
Federal Reserve's Check 21 products increased to 77.8 percent and 54.4
percent, respectively.
As a result of the declining check volumes, the Federal Reserve
developed a long-term plan for restructuring its check processing
operations. In 2003, the Federal Reserve had 45 check offices. Since
then, the Federal Reserve has closed a number of offices or gradually
eliminated its check processing operations. In June 2007, the Federal
Reserve announced that its check services system would be consolidated
into four regional check processing sites. As of September 30, 2008,
the Federal Reserve had 15 check offices and was working toward the
objective of maintaining four offices at Atlanta, Cleveland, Dallas,
and Philadelphia by the end of the first quarter of 2010. Given the
significant declines in paper check deposit volumes, the Federal
Reserve's Retail Payments Office believes that the Federal Reserve
likely will accelerate the consolidation schedule even further,
reducing its check processing offices to perhaps one office by mid-
2010.
Check Truncation Has Not Resulted Yet in Overall Gains in Economic
Efficiency for the Federal Reserve or for Banks, but Is Expected to
Produce Efficiencies in the Future:
Check truncation has not resulted yet in overall gains in economic
efficiency for the Federal Reserve or for the banks we surveyed, but
Federal Reserve and bank officials expect efficiencies in the future.
The expectation for electronic processing of checks was that it would
lead to gains in economic efficiency--that is, removing paper from the
payment stream would lead to lower costs. Our analysis of Federal
Reserve cost accounting data suggests that its costs may have increased
since the passage of Check 21, which may reflect concurrent maintenance
of its paper processing infrastructure, investments in equipment and
software for electronic check processing, and incurred costs associated
with closing check processing sites. Estimates varied on whether costs
were lower for private banks as the result of the check truncation that
Check 21 facilitated, reflecting differences in the ways in which
different banks handle checks and payments and differences among cost
accounting systems. For example, several of the 10 largest banks noted
that maintaining a dual paper-electronic infrastructure to date had
prevented them from achieving overall lower costs, although they had
seen reduced transportation and labor costs. Check imaging and the use
of substitute checks appear to have had a neutral impact on banks'
fraud losses.
Federal Reserve Cost Accounting Data Do Not Indicate Gains in Economic
Efficiency to Date, Partly Due to Maintenance of Dual Paper and
Electronic Processing Systems:
We found and the Federal Reserve's budget documents report that check
truncation has not decreased Federal Reserve costs, although it
contributed to decreased labor hours and transportation costs in
Federal Reserve check services. To distinguish the effects of check
truncation from other factors influencing the Federal Reserve's total
costs for check clearing services, we modified econometric cost
functions that Federal Reserve economists have used to assess the
effects of check volumes on total costs. In particular, we sought to
distinguish the effect of the increased use of check truncation
following passage of Check 21 on total costs from the concurrent
effects of:
* the decrease in the number of checks written in the United States,
* changes in the volume of checks processed by the Federal Reserve,
* the Federal Reserve's consolidation of its check services, and:
* costs of labor, software, and other expenses associated with the
check processing services.
With this consolidation of check offices, the Federal Reserve has
incurred an estimated $115 million in costs from 2003 through 2007,
including severance and other payments, which would increase total
check services costs. However, the Federal Reserve did recover all
costs for its check services from 2005 through 2007.[Footnote 15]
Consistent with our results, the Federal Reserve's annual budget
reports from 2006 through 2008 reported that the Federal Reserve's
budget for check services experienced cost overruns. Most recently, the
2008 annual budget review reported that the expense overrun was due
mainly to greater systemwide costs in preparation for additional
restructuring of check services (costs included $34.0 million for
accrual of severance, equipment impairments, and other
expenses).[Footnote 16] The 2007 annual budget review noted total
expenses for check services were to increase to $11.0 million
reflecting higher costs for Check 21-related supplies and equipment, as
well as additional resources necessary to facilitate further
consolidation into five regional check-adjustments sites.[Footnote 17]
Similarly, the 2006 annual budget review reported:
"Total check service expenses were budgeted to increase by $5.7
million, or 0.9 percent from the 2005 estimate. The increase reflects
one-time costs to prepare further consolidations of check operations,
as well as other initiatives underway to improve the efficiency of
check operations, including investments in Check 21 technology to
accommodate increased volumes."[Footnote 18]
The Planning and Control System (PACS) is the Federal Reserve's cost
accounting system for recording expenses, which includes the costs of
its check operations. We analyzed PACS data on check processing to
determine whether electronic check processing had an effect on total
processing costs. Our analysis builds on previous research by
economists in the Federal Reserve.[Footnote 19] The analysis includes
estimation of econometric cost functions using quarterly data from
first quarter of 1994 through the fourth quarter of 2007. We chose 1994
as the beginning point for the analysis based on conversations with
Federal Reserve officials about the data and in order to provide
adequate coverage for the period before and after enactment of Check
21. These cost functions estimate the effects that different
explanatory variables may have on total Federal Reserve costs for check
services. Explanatory variables include the total volume of checks
processed, the introduction of electronic processing or the volume of
checks processed electronically, the number of return items, the number
of Federal Reserve check processing offices, whether Check 21 was in
effect, and wage and price indexes. The cost functions permit isolation
of the effect of Check 21 from the effects of other variables on the
Federal Reserve's total costs for check services.
The results do not demonstrate any gains in economic efficiency as
measured by lower costs in the Federal Reserve's check operations for
the period since the passage of Check 21 through 2007. In particular,
the variable that would measure a change in total costs following the
effective date of Check 21 did not have a statistically significant
effect on total costs.[Footnote 20] See appendix II for a more detailed
discussion of the estimated cost functions. In part, the results
reflect costs associated with the concurrent closing of the Federal
Reserve's check processing sites. While these closings should reduce
costs in the long run, restructuring expenses incurred as part of the
closings (such as severance pay for workers) represent up-front costs.
The need to maintain dual infrastructures for paper and electronic
check services also may explain the results. While Check 21 removed a
barrier to electronic processing by creating the substitute check,
Check 21 did not require that paper be removed from the process. So,
the Federal Reserve continues to process paper checks and must maintain
the infrastructure to process paper checks as it invests in new
equipment to electronically process checks. Further, the creation of
the substitute check also required investment in new equipment to print
those instruments. For instance, a Federal Reserve Retail Payment
Office official noted that the high-speed printing machines for
substitute checks cost approximately $200,000 each and the Atlanta
processing site had purchased about 12 of these machines.[Footnote 21]
Although the move to electronic check services apparently has not led
yet to overall cost savings, the Federal Reserve has seen decreases in
transportation costs and work hours. With reduced paper volumes
accompanying check truncation, the Federal Reserve's transportation
costs for check services decreased approximately 11 percent from the
fourth quarter of 2001 through the fourth quarter of 2007 (see fig. 6).
Figure 6: Federal Reserve's Transportation Costs for Check Services,
1994-2007:
This figure is a combination line graph showing the Federal Reserve's
transportation costs for check services, 1994-2007. The X axis
represents the year and quarter, and the Y axis represents the dollars
in millions.
[See PDF for image]
Source: GAO analysis.
[End of figure]
The Federal Reserve also has seen a decrease in the number of work
hours for check services. Total work hours dropped from 2.6 million in
the fourth quarter of 2001 to 1.3 million in the fourth quarter of
2007, a decrease of approximately 48 percent (see fig. 7).
Figure 7: Federal Reserve Total Work Hours in Check Services, 1994-
2007:
This figure is a line graph showing the Federal Reserve's total work
hours in check services, 1994-2007. The X axis represents the year and
quarter, and the Y axis represents the work hours (in millions).
[See PDF for image]
Source: GAO analysis.
[End of figure]
The Largest U.S. Banks Still Maintain Dual Paper and Image-Based Check
Processing Systems for Check Collection and Noted Issues Affecting
Costs and Implementation:
Since the transition to imaging has been gradual throughout the banking
industry, the 10 largest U.S. banks still are maintaining paper-based
processing systems. As previously noted, Check 21 did not require banks
to take any action other than the acceptance of the substitute check.
The 10 largest banks in the United States, based on deposit size,
generally have large national branch networks and process large volumes
of checks; consequently, they have a financial incentive to reduce the
amount of paper they have to sort and transport. In 2007, these banks
individually had at least 350 million paper checks deposited by their
customers and some of them had considerably higher deposits, up to
approximately 5 to 7 billion checks.
But, the 10 banks have achieved various levels of electronic
processing. Two of the 10 banks have not converted their check
processing systems to imaging, but plan to do so by early 2009 and 7
banks have migrated to check imaging to some extent, but with imaging
volumes at various levels. As of 2007, on the basis of our data
collection instrument, the check volume of the seven banks that sent
electronic check images ranged from almost 4 to 60 percent of their
overall check deposits, although imaged volumes have been growing for
some of the seven banks. However, the seven imaging banks are
maintaining dual processing systems to collect on checks deposited at
their institutions. If a bank cannot receive an image, a bank or an
intermediary must either print a substitute check of the image or
present the original paper check.
Officials from four banks provided us with information on how the
continued use of paper presentment has affected their transition to
check imaging and their level of cost savings. Federal Reserve
officials noted that the willingness of private banks to invest in the
equipment needed to process check electronically demonstrated the
bank's expectation of lower costs. One bank official told us that the
bank still has to print substitute checks for presentment to the small
institutions that cannot receive images, which adds to the bank's
costs. Another bank noted that for banks that would prefer to receive
only paper, it will deposit the image with either the Federal Reserve
or another intermediary that then will print the substitute check to
present for payment. An official representing this bank stated that the
bank has to incur the additional cost of printing a substitute check
or, if it goes through an intermediary, to pay the intermediary's
prices. The same bank official added that maintaining paper operations
has delayed the ultimate potential savings from electronic check
processing because the bank had to keep in place its transportation
network to continue delivering paper checks. A third bank official
reported to us that fees paid to clear checks would be reduced as more
and more banks converted to imaging. Finally, a bank official from the
fourth bank advised us that mid-size and regional banks were behind in
their conversion to imaging because they are too large to outsource
their check business, but not large enough to have a financial
incentive to invest in check imaging technology. Thus, they continued
to use local clearinghouses where they could exchange their checks at
very low costs. This official noted that these banks need a reasonable
business case for investing in check imaging.
The declining volumes of paper checks also may be inhibiting the
migration of some banks to check imaging. As previously noted, from
2003 through 2006, the number of checks paid had declined from about 37
billion to over 30 billion checks. According to one bank trade
association, some banks are still undecided about converting to imaging
because they recognize that check volume is declining and wonder why
they should invest in check processing technology. During our
interviews, some of the seven imaging banks raised the issue of
declining check volumes as an additional complication preventing some
banks from converting to check imaging. Officials from the Federal
Reserve acknowledged while the volume of checks is declining, paper
checks would continue to be used long enough to warrant banks'
investments in the technology for a more efficient check processing
method. In both the paper-based and the image-based check processing
systems, the bank of first deposit bears most of the cost of check
collection; thus, it has the most financial incentive to convert to an
image-based system. In addition, under EFAA, the bank of first deposit
is required to release funds to the depositor within specified time
periods; thus, it has an additional incentive for speeding up
processing.[Footnote 22] The paying bank has the least market incentive
to migrate to imaging because it does not incur the costs for
collection, such as transportation and clearing fees.
Officials representing some of the four banks with the highest volumes
of check image deposits and receipts raised concerns with us that some
banks are refusing to migrate to the new imaging technology and some
action may be needed to encourage them to do so. One official told us
that paying banks should be paying more of the cost of check processing
so that they would have a financial incentive to receive images. The
official specifically stated that a group of banks has refused to
implement the technology and accept images. Another bank official said
that from approximately 5 to 7 percent of banks have refused to convert
to imaging and may need regulatory pressure to adopt the technology.
Largest Banks That Migrated to Imaging Achieved Cost Savings in the
Areas of Transportation and Labor, but also Incurred Technology Costs:
Under a paper-based check system, paper checks have to be sorted and
transported at every step until they are presented to paying banks; as
a result, transportation and labor are among the banks' highest costs.
From our analysis of responses to our data collection instrument,
officials from largest banks told us that labor was their largest
category of expenditures related to check processing followed by
transportation. However, none of the seven banks that process checks
electronically expect transportation to be a large expenditure category
for future processing operations if imaging technology is fully
implemented.
According to our bank interviews, air transportation networks of some
of the largest U.S. banks have been reduced. Four banks (those with the
highest volumes of check image deposits and receipts) have reduced
intrabank and interbank transportation routes for checks, particularly
air routes. By the end of 2009, two of the four will have eliminated
their air transportation networks entirely. However, three of the four
banks have not reduced costs for couriers and local transportation to
the same extent as for air transportation because they still transport
paper to central processing offices or to local clearinghouses.
We were told by two bank officials we interviewed that as more paper
checks are imaged at the branch level, the ground transportation costs
of banks should be reduced. One bank official advised us that the
earlier the bank can transmit the check information to its processing
system and capture the checks as images, the lower the bank's costs.
The official added that the bank is working toward implementing branch
"capture" (that is, conversion to an image) because the institution
achieves better float management and eliminates courier transportation
from its cost equation.[Footnote 23] Another bank official told us that
because his bank's transportation costs (for paper checks going from
the branches to the central processing office) would not be reduced
until the branches could capture check images; the bank had developed a
pilot program for capture in a few branches. Although imaging was
expected to result in savings in labor and transportation, the costs
associated with installing and maintaining imaging equipment and the
need to continue to maintain paper processing and clearing capabilities
has prevented the realization of cost savings. According to a third
bank, it is unclear when it will recover its significant investment in
imaging equipment, image archives, and image exchange enhancements, if
ever, due in part to the absence of universal adoption of check
imaging.
In contrast, we were told that transportation costs for banks that have
not migrated to electronic processing may increase because as the
overall volume of paper checks declines (due to check imaging and
consumer preference) transporting the remaining checks will become more
expensive on a per check basis. According to Federal Reserve officials,
when fewer banks require the services of a particular transportation
network, per-check transportation costs will increase for those banks
still using the services because the network is transporting a smaller
number of checks. The costs for the last bank on a specific route will
be very expensive. According to one Federal Reserve official, in the
future overnight mail may be the only practical option for these banks.
In congressional testimony, the Director of the Federal Reserve Board's
Division of Reserve Bank Operations and Payment Systems stated, "As
banks improve their technological capabilities, they can reduce their
reliance on air and ground transportation, especially shared
transportation arrangements. The banks that remain tied to paper checks
will continue to bear the costs of those arrangements."[Footnote 24]
Furthermore, bank officials told us that they had additional technology
costs when they converted to a check imaging system. To exchange checks
electronically with other banks, banks needed to adapt their systems
both to send and receive images. The technologies required for
electronic check processing include hardware and software to image
checks, archive images, and transmit image cash letters for collection.
From the analysis of responses to our data collection instrument, six
banks projected that the technology costs would continue to be in the
"great" or "greatest" range for the foreseeable future. On the basis of
our interviews, the two largest imaging banks have recovered or will
recover the investments they made for check imaging by 2009. An
official representing one of the three banks stated that the bank
recovered its investment in imaging mostly through savings in labor and
transportation. Moreover, the bank had less equipment, lower
maintenance costs on the remaining equipment, and needed less back
office space because of electronic processing. The banks that have not
recovered their investments still were investing in image archive and
image exchange enhancements.
Similar to the Federal Reserve, banks have to deal with substitute
checks and, thus, may be required to invest in the printing of
substitute checks. From the analysis of responses to our data
collection instrument, officials representing banks that have deposited
images categorized expenditures for the printing of substitute checks
in the "some" to "very great" range. In a follow-up interview, one bank
official told us that the bank decided to outsource the printing
because it decided not to make the investment since substitute checks
were a temporary measure and would not be used once all institutions
were image-enabled. Thus, this investment did not make sense for the
bank. Another bank official acknowledged that substitute check printing
has cost the bank hundreds of thousands of dollars to implement.
A Few Smaller-Size Banks Have Seen Lower Costs for Transportation and
Labor from Electronic Check Processing:
Smaller banks also have been migrating to electronic check processing.
But, according to our interviews with three smaller banks (in this
case, one bank and two credit unions), they have migrated all of their
volumes to electronic processing rather than operating two processing
systems, as the largest banks have been doing. In addition, the three
smaller banks told us that they typically will use a third-party
processor, an image exchange processor like Endpoint Exchange, the
Federal Reserve, or another intermediary, such as a correspondent bank.
For example, a credit union deposited and received images through the
Federal Reserve Banks, while a medium-size bank, with assets of $4.4
billion, deposited and received images through an image processor and
correspondent.
Officials representing the smaller banks told us that it may be easier
for small banks to completely migrate to imaging because their check
volumes are minuscule in comparison to the volumes of the largest banks
and their back offices generally are less complicated than those of the
largest banks. The bank with $4.4 billion in assets received
approximately 15 million checks for deposit in 2007, compared with the
10 largest banks in which the bank with the lowest volume of check
deposits had 350 million checks deposited. Moreover, generally when
these institutions migrate to check imaging, they acquire the imaging
services of their intermediary or processor rather than creating their
own.
In our interviews, representatives of the smaller banks described how
check imaging had affected their operations and costs. The bank with
assets of $4.4 billion reduced its costs by reducing its transportation
network. According to a bank official, the bank also expects to secure
cost savings from its local courier routes in the future. But, the bank
had to invest in software to transfer check images to its correspondent
bank. An official from a small credit union told us that check imaging
allowed it to reduce its labor costs by half, after spending almost
$6,000 for technology. Another credit union told us that they were able
to eliminate three full-time equivalent positions because check
processing and related operations (such as researching customer issues
on payments) became more efficient. According to an official at the
credit union, while the institution made some investments in technology
and software, it had recovered the investment costs because of the
staff reductions.
Use of Substitute Checks and Check Imaging Appears to Have Had a
Neutral Effect on Fraud Losses:
Based on a recent American Bankers Association's (ABA) survey of their
members about fraud in deposit accounts, the analysis of responses to
our data collection instrument, and our interviews with banks, we found
that the use of substitute checks and check imaging has had a neutral
effect on fraud losses.[Footnote 25] In 2007, the ABA reported in its
survey of members, more than 92 percent of the bank respondents
answered that they had not incurred any losses from substitute checks
in 2006. Of the 8 percent of banks that responded that they had
incurred both fraud and non-fraud losses from substitute checks, more
than 80 percent also responded that these losses did not occur because
the instruments were substitute checks instead of original checks.
From the analysis of our responses to our data collection instrument,
the six largest banks that have migrated to electronic check processing
noted that check imaging and the use of substitute checks had not
affected the prevalence of losses from bad checks and that imaging has
had a neutral or minimal effect on check fraud. Officials representing
two of these banks explained in subsequent interviews that in the post-
Check 21 world, since checks are being processed faster banks can catch
a fraudulent item sooner. A third official told us that he had seen a
slight decline in fraud losses since Check 21. Finally, from the
analysis of the responses to our data collection instrument, four of
the largest banks noted that they had not taken additional actions to
alleviate the potential threat of losses from images of bad checks.
Most Bank Consumers Appeared to Have Accepted Changes to Their Checking
Accounts from the Check Truncation Process Resulting from Check 21:
On the basis of our structured bank consumer interviews, we found only
a small percentage of consumers who preferred to receive canceled
checks with their checking account statement. Of the bank consumers we
interviewed, 12 (or about 11 percent) wanted their canceled checks
returned, while 37 (or about 35 percent) preferred to use online
banking capabilities to review their check payment activity. In
general, consumers expressed a variety of preferences for how banks
should provide them with the most complete information about their
check payments activity. Also, most of the consumers were not concerned
significantly about being able to demonstrate proof of payment using a
substitute check or check image rather than a canceled check. Few of
the consumers reported that they suffered errors from the check
truncation process. In addition to conducting consumer interviews, we
reviewed consumer complaint data provided by federal banking regulators
and found relatively few consumer complaints relating to Check 21.
A Small Percentage of Bank Consumers Preferred Receiving Their Canceled
Checks:
We found that a small percentage of bank consumers in our structured
interviews preferred receiving canceled checks, while the remaining
consumers preferred reviewing their check payments activity online or
in a less paper-intensive format, such as image statements. As we
reported in an earlier report, perceptions about consumer preferences
for the receipt of their canceled checks deterred the adoption of
electronic check processing.[Footnote 26] Based on the bank consumers
we interviewed, it appears that their preference for canceled checks is
diminishing. In our interviews, consumers expressed a variety of
preferences for how banks should provide them with the most complete
information about their check payments activity (see fig. 8).
Figure 8: Bank Consumer Preferences for Reviewing Check Payments
Activity:
This figure is a pie graph showing bank consumer preferences for
reviewing check payments activity.
On-line: 35%;
Combination of check images and on-line: 16%;
Paper checks: 11%;
Check images: 10%;
Combination of paper checks and on-line review: 7%;
Combination of paper checks and check images: 7%;
Combination of substitute checks and on-line: 5%;
Combination of substitute checks and check images: 3%;
Combination of all: 2%;
Substitute checks: 2%;
Combination of paper checks, substitute checks, and on-line review: 2%;
Combination of substitute checks and paper checks: 1%.
[See PDF for image]
Note: Percentages are based on responses from 107 bank consumers.
[End of figure]
In particular, 12 of the 107 consumers, or about 11 percent, told us
that they preferred receiving their canceled checks with their checking
account statement.[Footnote 27] Some of these consumers believed that
canceled checks were better for recordkeeping and more secure than
electronic images in terms of protecting their privacy. Others in this
group stated they wanted to be able to review their handwriting and
other details of the canceled paper check to ensure that the checks
were not counterfeit or the signatures forged. However, most bank
consumers we interviewed accepted the use of online banking to review
their check payments activity. Specifically, 37 of the 107 consumers,
or about 35 percent, told us that they preferred reviewing check
information and images online.[Footnote 28] Several consumers stated
that they did not need the "extra paper" from canceled checks and image
statements and that online reviewing was more secure than receiving
canceled checks. Some consumers stated that they enjoyed the
convenience of reviewing their check payments activity online at any
time. Twenty-eight of the 107 consumers, or 26 percent, preferred a
combination of the various methods (check images, online review, paper
checks, and substitute checks).
Most Bank Consumers Were Not Concerned Significantly about
Demonstrating Proof of Payment Using a Substitute Check or Check Image:
Most bank consumers reported that they were not concerned significantly
about demonstrating proof of payment despite the changes to their
checking accounts resulting from check truncation. For example, a
consumer might pay a debt using a check, but the creditor might not
properly record the payment, and then ask the consumer to demonstrate
proof that he or she paid. Under the check truncation process, the
consumer most likely would have access only to a substitute check or an
image of the canceled check and not the original, canceled check.
In our structured interviews, we asked consumers about their experience
with demonstrating proof of payment. We found that 33 of the 108
consumers, or about 31 percent, had never been required to demonstrate
proof of payment using canceled checks, substitute checks, or an image
statement. We found that 58 of the 108 consumers, or about 54 percent,
had used a canceled check to demonstrate proof of payment. We also
found that 33 of the 108, or about 31 percent, had used a substitute
check or image statement to demonstrate proof of payment. Most of these
consumers reported that they had no difficulty using a substitute check
or image statement, but some consumers reported that creditors would
not accept an image showing only the front of the check so the consumer
had to get copies of the front and back of the check from the bank.
We then asked consumers whether they were concerned about having to
demonstrate proof of payment using a substitute check or image
statement rather than a canceled check. We found that 53 of the
consumers, or about 49 percent, were "slightly" or "not at all"
concerned about their ability to demonstrate proof of payment using a
substitute check or image statement (see fig. 9). In particular, many
of these consumers were confident that a substitute check or image
statement contained all of the information necessary to demonstrate
proof of payment. However, 35 of the consumers, or 32 percent, were
"extremely" or "very" concerned about using a substitute check or image
statement. Many of these consumers were concerned that having an image
of only the front of the check might not be sufficient, particularly if
they had experienced such difficulty in the past.
Figure 9: Bank Consumer Concern about Demonstrating Proof of Payment
Using a Substitute Check or Image Statement:
This figure is a pie graph showing bank consumer concern about
demonstrating proof of payment using a substitute check or image
statements.
Not concerned at all: 27%;
Slightly concerned: 22%;
Moderately concerned: 19%;
Very concerned: 18%;
Extremely concerned: 15%.
[See PDF for image]
Source: GAO.
Note: Percentages may not total to 100 percent due to rounding.
[End of figure]
Few Bank Consumers Reported That They Experienced Errors from the Check
Truncation Process:
Few of the bank consumers we interviewed reported that they suffered
errors from the check truncation process. We asked consumers whether
they had experienced errors such as double-posting of an item, a forged
signature on a check, a counterfeit check, or some other error
involving canceled checks, substitute checks, and image
statements.[Footnote 29] The consumers reported more errors involving
canceled checks than substitute checks or image statements.
Specifically, 28 of the 108 consumers, or about 26 percent, reported an
error involving a canceled check and using it to resolve the error. In
contrast, only one consumer we interviewed reported suffering an error
related to double-posting of a debit and using a substitute check to
resolve the error. Also, 7 of the 74 consumers who reported that they
received image statements, or about 9 percent, reported errors
involving an image statement and using it to resolve errors they
experienced. See figure 10 for the distribution of reported errors
involving canceled checks and image statements.
Figure 10: Errors Reported by Bank Consumers Involving Canceled Checks
and Image Statements:
This figure is a combination of two pie graphs showing errors reported
by bank consumers involving canceled checks and image statements.
Have you experienced any of the following potential errors in involving
your canceled check?
None: 74%;
Double-posting of an item: 11%;
Other: 9%;
Forged signature on one of your checks: 5%;
Counterfeit check: 1%.
Have you experienced any of the following potential errors involving in
a check image?
None: 91%;
Double-posting of an item: 4%;
Other: 1%;
Forged signature on one of your checks: 3%;
Counterfeit check: 1%.
[See PDF for image]
Source: GAO.
Note: Percentages in the first graphic are based on responses from 108
bank consumers. Percentages in the second graphic are based on
responses from 74 bank consumers.
[End of figure]
Based on interviews with trade association and service vendor
officials, we found that some banks have been correcting errors
associated with double-posting of a check before consumers experience
them. They told us that double-posting initially was a significant
problem for banks as they adopted check truncation technology. However,
they also noted that many banks have now incorporated protection in
their computer system to identify duplicates before they reach the
consumer, so that many consumers never see them when they review their
bank statements.
Federal Banking Regulators Reported Few Consumer Complaints on Check
21:
We found that a small percentage of consumers complained to the federal
banking regulators about matters relating to Check 21.[Footnote 30] In
its April 2007 report, the Federal Reserve Board found that less than 1
percent of all complaints received by federal banking regulators
related to Check 21. The results of our review of consumer complaint
data on Check 21 corroborated the Federal Reserve Board's conclusion.
Specifically, we reviewed consumer complaint data from the four federal
banking regulators from October 28, 2004, through March 31, 2008, and
found 172 complaints were submitted about Check 21. In comparison, in
each year from 2005 through 2007, the regulators received approximately
35,000 consumer complaints overall. Of the 172 complaints relating to
Check 21, we found that 78, or about 45 percent, were from consumers
who wanted to continue receiving canceled checks. The federal banking
regulators responded to such complaints by noting that banks have no
legal requirement to return canceled checks to consumers and that the
return of canceled checks was dependent on the contractual agreement
between consumers and their banks. However, in these instances, the
data showed that the interested banks generally agreed to send canceled
checks to consumers whenever possible. In addition, another 30 of the
172 complaints, or about 17 percent, were from consumers concerned
about the quality or clarity of image statements. Some of the banks we
interviewed also mentioned image quality as a prominent consumer
complaint, but we learned that they continue to seek a solution to
image quality problems.
Bank Consumers Have Realized Benefits and Costs Relating to Faster
Check Processing and Access to Information about Their Checking
Accounts:
To the extent that banks have implemented electronic check processing,
bank consumers have realized both benefits and costs relating to faster
processing and access to information about their checking accounts.
Faster check processing has helped some banks extend the cut-off time
for same-day credit on deposits, which can result in faster
availability of deposited funds. In addition, bank industry officials
and some of the consumers we interviewed believe it is beneficial to
receive simpler checking account statements with check images rather
than canceled checks. Also, bank industry officials cited benefits to
consumers from immediate access to information about checking account
activity and improved customer service. In addition, consumers can
benefit specifically from a provision of Check 21 because they have the
right to expedited re-credit of their checking accounts if banks make
certain errors associated with substitute checks. However, on the basis
of our consumer and bank interviews, the extent to which consumers have
benefited from expedited re-credit is unclear. We also found that some
consumers may incur fees related to receiving canceled checks and check
images with their checking account statements. Based on our review of
available data from 2001 through 2006, it appears that fees for
canceled checks have increased and fees for check images have remained
relatively flat. In addition, the amount of the fees can vary depending
on the type of checking account the consumer maintains.
Some Bank Consumers May Experience Extended Cut-off Time for Deposits,
but Broader Adoption of Electronic Processing by Banks Would Be Needed
to Shorten Funds Availability Deadlines:
We found that banks may have extended the cut-off time for accepting
deposits for credit on the same business day, due to the check
truncation process and other check-system improvements. Generally,
banks had established a cut-off hour of 2:00 p.m. or later for receipt
of deposits at their main or branch offices and a cut-off of 12:00 p.m.
or later for deposits made at ATMs and other off-premise
facilities.[Footnote 31] These cut-off times provided the banks with
necessary time for handling checks and transporting them overnight to
paying banks. The check truncation process and check imaging provide
collecting banks with additional time to present checks to paying
banks. As a result, banks may be able to establish a later cut-off
hour, which would give consumers more time to deposit funds at the bank
for same-day credit.
Bank officials told us that they have started to adjust their cut-off
times in some geographic areas in response to the growth of check
truncation. Of the seven largest U.S. banks that have started to
migrate to check imaging, five told us that they have extended some of
their deposit cut-off times at certain branches. For instance, one bank
on average extended its cut-off time by 2 hours in the Northeast, and
another bank had plans in place to make a similar 2-hour extension in
selected markets. A third bank told us that it has extended the cut-off
time for accepting deposits for credit on the same business day at
certain ATMs to 8:00 p.m. in several major cities such as Atlanta,
Chicago, Los Angeles, and New York.
Although some consumers may have additional time for making deposits,
they may not be able to withdraw their funds any sooner because the
funds availability schedules of Regulation CC have not been amended
following enactment of Check 21. The Federal Reserve Board recently
concluded that much broader adoption of new technologies and processes
by the banking industry must occur before check return times can
decline appreciably and thereby permit a modification of the funds
availability deadlines.[Footnote 32] The Federal Reserve Board found
that the banks of first deposit learn of the nonpayment of checks
faster than they did when EFAA was enacted, but banks still do not
receive "most" local or nonlocal checks before they must make funds
available for withdrawal.[Footnote 33]
However, the Federal Reserve's decision to consolidate its check-
processing regions has had a direct effect on consumers in terms of the
availability of their deposited funds under Regulation CC.
Specifically, the consolidations have increased the proportion of local
checks and thereby reduced the maximum permissible hold period from 5
business days to 2 business days for many checks. As previously noted,
the Federal Reserve's check-processing regions are being consolidated
into four check-processing regions by the first quarter of 2010.
Because the processing regions are larger (and will become even more
so), the number of local checks has been increasing.
In addition, based on the Federal Reserve Board's study and our own
research, it appears that banks are making depositor funds available
earlier than EFAA-established funds-availability schedules.
Specifically, the Federal Reserve's Check 21 study found that banks
make about 90 percent of all consumer deposits of local and nonlocal
checks available more promptly than required by EFAA.[Footnote 34]
Moreover, it found that banks make funds available from the majority of
consumer check deposits within 1 business day.[Footnote 35] We reviewed
the customer account agreements for 5 of the 10 largest U.S. banks and
found that the general policy for each bank is to make funds available
to consumers on the business day after the day of deposit.
Many Bank Consumers Have Realized Other Benefits Related to Access to
Information about Check Payments:
Bank industry officials and some consumers we interviewed noted that
consumers may realize other benefits relating to access to information
about check payments. For example, bank consumers may receive simpler
checking account statements using image technology. So-called "image
statements" include a sheet of paper with multiple pictures or images
of checks that were written by the consumer and processed since the
last statement. In our interviews with 108 bank consumers, 75
consumers, or about 69 percent, stated that they received image
statements. When asked about their preferred method of receiving
information about check payments, 11 of the 108 consumers interviewed,
or about 10 percent, stated that they preferred receiving image
statements over canceled checks or online review of check payments
activity. Some of the 11 consumers told us that they preferred
receiving image statements because, while they wanted a paper record of
their check payments activity, they preferred not to handle and store
canceled checks.
Bank consumers who prefer to manage their checking account
electronically also might realize benefits from immediate access to
information about check payments. With the check imaging process and
online access to their checking accounts, consumers can review check
payments and images of their paid checks as soon as they are posted to
the account and may recognize a problem sooner. With paper check
processing, consumers must wait until the checking account statement
arrives in the mail to review their check payments activity. Also,
improved access to information can be beneficial to consumers when they
need to work with the bank to resolve a problem. Bank industry
officials and some consumers we interviewed noted that consumers may
realize other benefits relating to access to information about check
payments.
Check 21's Expedited Recredit Is Considered a Consumer Benefit, but It
Appears That a Limited Number of Consumers Filed for the Benefit:
One of the expected consumer benefits of Check 21 is the right to
expedited recredit, but the extent to which consumers have benefited is
unclear. The expedited recredit provision is considered a benefit to
consumers because other banking laws governing checks do not prescribe
specific amounts or time frames by which banks must recredit a
customer's account.[Footnote 36] On the basis of our bank consumer and
bank interviews, it appears that a small number of bank consumers have
filed expedited recredit claims. The right to expedited recredit exists
if the consumer asserts in good faith that the bank charged the
consumer's account for a substitute check provided to the consumer and
either the check was not properly charged to the consumer's account, or
the consumer has a warranty claim pertaining to the substitute
check.[Footnote 37] The bank must recredit the customer's account
unless it has provided the customer the original check or a copy of the
original check that accurately represents all information on the
original check and demonstrated to the consumer that the substitute
check was properly charged to the consumer's account.[Footnote 38]
On the basis of our consumer and bank interviews, it appears that a
small number of bank consumers have filed expedited recredit claims. In
our interviews with 108 consumers, 9 or about 8 percent of the
consumers we interviewed, stated that they had received substitute
checks with their main checking account statement, and none had
exercised the right to expedited recredit. On the basis of the data
provided to us by the 10 largest banks through the data collection
instrument (which are not representative of the entire industry), we
found 3 banks received a small number of claims related to expedited
recredit in 2007. Specifically, one bank reported that it fielded less
than 1,000 claims; one received less than 10 claims; and the third bank
reported that it received 1 claim.[Footnote 39] In an interview, a
representative of another bank told us that the bank had not received
any claims. Six other banks did not report any information on the
number of claims received.
Some Bank Consumers Can Incur Fees for Receiving Canceled Checks and
Image Statements:
Some bank consumers can incur fees for receiving canceled checks and
image statements, and the amount can depend on the type of checking
account the consumer maintains. We reviewed data regarding bank fees
for canceled checks and image statements acquired from Informa Research
Services in conjunction with a report on bank fees.[Footnote 40] The
data indicated that the average amount of fees for obtaining canceled
checks generally increased from 2001 through 2006, and the average
amount of fees for obtaining image statements remained relatively flat.
For example, as shown in figure 11, the average check enclosure fee
more than doubled from $1.42 to $3.11. During the same period, the
average check imaging fee rose from $0.40 to $0.49.[Footnote 41]
Figure 11: Check Enclosure and Imaging Fees, Banks and Savings and
Loans, 2001-2006:
This figure is a combination line graph showing check enclosure and
imaging fees, banks and savings loans, 2001-2006. The X axis represents
the year, and the Y axis represents the average fee (dollars).
Year: 2001;
Check enclosure fee: 1.42;
Check imaging fee: 0.4.
Year: 2002;
Check enclosure fee: 1.72;
Check imaging fee: 0.38.
Year: 2003;
Check enclosure fee: 1.77;
Check imaging fee: 0.33.
Year: 2004;
Check enclosure fee: 2.14;
Check imaging fee: 0.39.
Year: 2005;
Check enclosure fee: 2.83;
Check imaging fee: 0.56.
Year: 2006;
Check enclosure fee: 3.11;
Check imaging fee: 0.49.
[See PDF for image]
Source: GAO analysis of Informa Research Service data.
[End of figure]
The Informa data also indicated that banks may charge different amounts
for check enclosures and check imaging depending on the type of
checking account. Specifically, the Informa data indicated that
primarily non-interest, free checking accounts had the highest fees for
check enclosures and check imaging.[Footnote 42] The lowest check
enclosure and check imaging fees were found primarily with senior
checking accounts.[Footnote 43] For example, in 2006 the average check
enclosure fees for a non-interest, free checking account and a senior
checking account were $3.75 and $2.45, respectively, compared to $3.11-
-the average check enclosure fee of all accounts Informa surveyed.
Furthermore, the average check-imaging fee for a non-interest, free
checking account in 2006 was $0.84, and the average check-imaging fee
for a senior checking account was $0.18, compared to $0.49--the average
check imaging fee of all accounts Informa surveyed.
A relatively small number of the bank consumers we interviewed reported
that their bank charged a fee for obtaining canceled checks or image
statements, and some of the banks we interviewed reported that they
charged a fee for providing canceled checks. Specifically, 23 bank
consumers, or about 21 percent of the consumers we interviewed, told us
that their bank charged a fee for obtaining canceled checks. Two
consumers stated that they switched to online review of their check
payments activity to avoid paying a fee for receiving canceled checks.
Also, as we reported above, 12 of the 108 bank consumers we interviewed
preferred receiving canceled checks to review their check payments
activity. Moreover, 18 bank consumers, or about 17 percent, reported
that their bank charged a fee for obtaining image statements. Two of
the banks we interviewed charged a fee if consumers wanted to receive
canceled checks. For example, one bank stated that its customers paid
$2 for receiving canceled checks if they also paid a monthly service
fee, but other bank officials we interviewed stated that their banks
did not charge a fee for image statements.
In addition, faster check processing may cause consumers to lose
"float." Float is the time between the payment transaction and the
debiting of funds from a bank consumer's account. The check truncation
process may result in checks clearing a consumer's account more quickly
than under traditional check processing. However, deposited funds may
not be available to consumers more quickly because, as noted above,
Regulation CC's funds availability deadlines have not changed.
According to our recent report on bank fees, consumer groups and bank
representatives believe that the potential exists for increased
incidences of overdrafts if funds were debited from a consumer's
account faster than deposits were made available for
withdrawal.[Footnote 44] However, we identified little research on the
extent to which check truncation has affected occurrences of overdrafts
and nonsufficient funds fees.[Footnote 45]
Agency Comments and Our Evaluation:
We provided a copy of a draft of this report to the Federal Reserve
Board, which provided us with written comments that are reprinted in
appendix III. The Federal Reserve Board agreed with our overall
conclusion that, over the past four years, the banking industry has
made substantial progress toward establishing an end-to-end electronic
check-processing environment. In commenting on this report, the Federal
Reserve Board noted that the Federal Reserve Banks expect that by year-
end 2009, more than 90 percent of their check deposits and presentments
will be electronic. They also commented that the ongoing transformation
to electronic check-processing environment has not been without cost.
As noted in our report, the Federal Reserve Banks have reduced their
transportation costs and work hours associated with their check
services. And, according to the Federal Reserve Board, they earned a
net income of $326 million for providing check services from 2005
through 2007. The Federal Reserve Board concurred with a number of
consumer benefits identified in the report: faster funds availability
on check deposits due to later deposit deadlines, quicker access to
account information, and improved customer service. In addition, they
provided us with technical comments, which we incorporated as
appropriate. We also sent a draft of this report to the Federal Deposit
Insurance Corporation, Office of the Comptroller of the Currency, and
Office of Thrift Supervision. Only the Office of the Comptroller of the
Currency provided us with technical comments, which we incorporated as
appropriate. We provided sections of the draft of this report to bank
officials for their technical review and several of them provided us
technical comments, which we incorporated as appropriate.
We are providing copies of this report to other interested
Congressional committees. We are also providing copies of this report
to the Chairman, Board of Governors of the Federal Reserve System;
Chairman, Federal Deposit Insurance Corporation; Comptroller of the
Currency, Office of the Comptroller of the Currency; Director, Office
of Thrift Supervision; and other interested parties. We will also make
copies available to others upon request. In addition, the report will
be available at no charge on the GAO Web site at [hyperlink,
http://www.gao.gov].
If you or your staffs have any questions regarding this report, please
contact me at (202) 512-8678 or jonesy@gao.gov. Contact points for our
Office of Congressional Relations and Public Affairs may be found on
the last page of this report. GAO staff who made major contributions to
this report are listed in appendix IV.
Signed by:
Yvonne D. Jones:
Director, Financial Markets and Community Investment:
List of Congressional Committees:
The Honorable Christopher J. Dodd:
Chairman:
The Honorable Richard C. Shelby:
Ranking Member:
Committee on Banking, Housing, and Urban Affairs:
United States Senate:
The Honorable Barney Frank:
Chairman:
The Honorable Spencer Bachus:
Ranking Member:
Committee on Financial Services:
House of Representatives:
The Honorable Carolyn B. Maloney:
Chair:
The Honorable Judy Biggert:
Ranking Member:
Subcommittee on Financial Institutions and Consumer Credit:
Committee on Financial Services:
House of Representatives:
[End of section]
Appendix I: Objectives, Scope, and Methodology:
The Check Clearing for the 21st Century Act of 2003 (Check 21) mandated
that GAO evaluate the implementation and administration of Check 21.
The report objectives are to: (1) determine the gains in economic
efficiency from check truncation and evaluate the costs and benefits to
banks and the Federal Reserve System (Federal Reserve) from check
truncation, (2) assess consumer acceptance of the check truncation
process resulting from Check 21, and (3) evaluate the costs and
benefits to consumers from check truncation.[Footnote 46]
To estimate the gains in economic efficiency from check truncation and
evaluate the costs and benefits to banks from check truncation, we
separately analyzed costs for the check operations of the Federal
Reserve and for a selected group of banks. We used data from the
Federal Reserve cost accounting system, known as the Planning and
Control System or PACS, for the period beginning 10 years prior to the
effective date of Check 21 (1994) through 2007. We modeled the Federal
Reserve's total check processing costs as different functions of
variables, such as the volume of checks processed, the volume of
returned checks, the number of Federal Reserve check processing
offices, and the general indexes on wage and price. The specified cost
functions allowed us to use standard econometric methods for estimating
the effects of the variables on the Federal Reserve's total check
processing costs for 1994 through 2007. Because data on prices of input
factors associated with Federal Reserve's check processing operations
are not available, we also used in our estimation data from the
Department of Commerce's Bureau of Economic Analysis (BEA) and the
Department of Labor's Bureau of Labor Statistics (BLS) as alternative
measurements for the prices of these input factors. For example, we
used average hourly earning for all private sectors from BLS as an
alternative measurement for the Federal Reserve's labor cost, BEA's
price deflator for equipment and software by nonresidential producers
as an alternative measurement for communications equipment and transit
cost, and BEA's Gross Domestic Product price deflator as an alternative
measurement for costs of all other input factors. We assessed the
quality of all the above data and found them to be sufficiently
reliable for our purposes. We also discussed Federal Reserve check
processing costs and our econometric cost model with staff at the
Federal Reserve. See appendix II for a detailed discussion of our
econometric cost functions.
While the Federal Reserve has consistent cost accounting data, cost
accounting varies throughout the banking industry, preventing a similar
analysis for private-sector costs. To evaluate the costs and benefits
to banks from check truncation, we focused our data collection and
analysis on the 10 largest banks in the United States, based on deposit
size as of March 25, 2008. The check volume at the 10 largest U.S.
banks represents a significant segment of the check paid volume. In
2007, these banks presented almost 13 billion checks for collection out
of approximately 30 billion checks, which were paid in 2006. Thus, we
determined that these banks should have a financial incentive to reduce
the amount of paper that has to be sorted and transported. We created a
data collection instrument to obtain qualitative cost information about
the following issues: (1) the extent to which the banks deposited and
received checks as images; (2) the primary costs related to paper check
processing; (3) the extent of the investment that banks made to
exchange check images; (4) the level of cost savings banks achieved, if
any, including changes in labor and transportation costs through the
use of image technology; and (5) the impact of check imaging and the
use of substitute checks on the prevalence of bank losses from
fraudulent checks. Officials from the Electronic Check Clearing House
Organization, commonly known as ECCHO, also reviewed the data
collection instrument. We sent it to the 10 banks and received a
response from 9. At an early stage of our engagement, we also
interviewed an official representing the bank that did not provide a
response. We conducted follow-up interviews with a number of the banks
requesting clarification of their responses.
We also sent the data collection instrument to 12 smaller institutions,
which included credit unions, to understand the small bank experience
with check imaging. These banks' assets ranged from less than $500
million to $5 billion and were selected from ECCHO's list of
participating members. In addition, our selection criteria included
whether these smaller institutions were located in metropolitan or
nonmetropolitan areas. We received completed forms from five of these
institutions, but two had not migrated any of their volume to check
imaging. We conducted subsequent interviews with the three institutions
that had. We made several attempts to contact the nonrespondents
through e-mail messages and follow-up telephone calls. In addition, we
interviewed officials from a corporate credit union and a banker's
bank.[Footnote 47]
To assess consumer acceptance of the check truncation process resulting
from Check 21, we conducted in-depth structured interviews with a total
of 108 adult consumers in three locations (Atlanta, Boston, and
Chicago) in May 2008. We contracted with NuStats, Inc., a private
research and consulting firm, to recruit a sample of consumers who
generally represented a range of demographics within the U.S.
population in terms of age, education level, and income. However, the
consumers recruited for the interviews did not form a random,
statistically representative sample of the U.S. population; therefore,
we could not generalize the results of the interviews to the relevant
total population. Additionally, the self-reported data we obtained from
consumers are based on their opinions and memories, which may be
subject to error and may not predict their future behavior. Consumers
had to speak English and meet certain other conditions: having primary
responsibility in the household for balancing the financial account
that allows paper check writing; having received canceled original
checks in paper form with the checking account statement at some point
since 2000; and not having participated in more than one focus group or
similar in-person study in the 12 months before the interview. We
achieved our sample recruitment goals for all demographics, with the
exception of the age category "65 plus" and the education category
"some high school or less." In addition, our sample comprised 64 women
and 43 men. We considered that the impact of not achieving these goals
on our work was minimal. See table 1 for further demographic
information on the consumers we interviewed.
Table 1: Demographic Characteristics of Consumers Interviewed:
Demographic: Income: Less than $25,000;
Atlanta: 9;
Atlanta: 23%;
Chicago: 7;
Chicago: 18%;
Boston: 5;
Boston: 17%;
Total: N: 21;
Total: %: 19%.
Demographic: Income: $25,000 to $44,999;
Atlanta: 7;
Atlanta: 18%;
Chicago: 10;
Chicago: 26%;
Boston: 4;
Boston: 14%;
Total: N: 21;
Total: %: 19%.
Demographic: Income: $45,000 to $64,999;
Atlanta: 10;
Atlanta: 25%;
Chicago: 8;
Chicago: 21%;
Boston: 7;
Boston: 24%;
Total: N: 25;
Total: %: 23%.
Demographic: Income: $65,000 to $100,000;
Atlanta: 9;
Atlanta: 23%;
Chicago: 10;
Chicago: 26%;
Boston: 5;
Boston: 17%;
Total: N: 24;
Total: %: 22%.
Demographic: Income: More than $100,000;
Atlanta: 5;
Atlanta: 13%;
Chicago: 4;
Chicago: 10%;
Boston: 8;
Boston: 28%;
Total: N: 17;
Total: %: 16%.
Demographic: Income: Total;
Atlanta: 40;
Atlanta: 100%;
Chicago: 39;
Chicago: 100%;
Boston: 29;
Boston: 100%;
Total: N: 108;
Total: %: 100%.
Demographic: Age: 25-34;
Atlanta: 7;
Atlanta: 18%;
Chicago: 9;
Chicago: 23%;
Boston: 7;
Boston: 24%;
Total: N: 23;
Total: %: 21%.
Demographic: Age: 35-44;
Atlanta: 10;
Atlanta: 25%;
Chicago: 9;
Chicago: 23%;
Boston: 6;
Boston: 21%;
Total: N: 25;
Total: %: 23%.
Demographic: Age: 45-54;
Atlanta: 9;
Atlanta: 23%;
Chicago: 9;
Chicago: 23%;
Boston: 6;
Boston: 21%;
Total: N: 24;
Total: %: 22%.
Demographic: Age: 55-64;
Atlanta: 8;
Atlanta: 20%;
Chicago: 9;
Chicago: 23%;
Boston: 6;
Boston: 21%;
Total: N: 23;
Total: %: 21%.
Demographic: Age: 65+;
Atlanta: 6;
Atlanta: 15%;
Chicago: 3;
Chicago: 8%;
Boston: 4;
Boston: 14%;
Total: N: 13;
Total: %: 12%.
Demographic: Age: Total;
Atlanta: 40;
Atlanta: 100%;
Chicago: 39;
Chicago: 100%;
Boston: 29;
Boston: 100%;
Total: N: 108;
Total: %: 100%.
Demographic: Education: Some high school or less;
Atlanta: 4;
Atlanta: 10%;
Chicago: 4;
Chicago: 10%;
Boston: 0;
Boston: 0;
Total: N: 8;
Total: %: 7%.
Demographic: Education: Completed high school;
Atlanta: 10;
Atlanta: 25%;
Chicago: 10;
Chicago: 26%;
Boston: 8;
Boston: 28%;
Total: N: 28;
Total: %: 26%.
Demographic: Education: Some college;
Atlanta: 10;
Atlanta: 25%;
Chicago: 11;
Chicago: 28%;
Boston: 9;
Boston: 31%;
Total: N: 30;
Total: %: 28%.
Demographic: Education: Completed college;
Atlanta: 11;
Atlanta: 28%;
Chicago: 7;
Chicago: 18%;
Boston: 8;
Boston: 28%;
Total: N: 26;
Total: %: 24%.
Demographic: Education: Graduate school;
Atlanta: 3;
Atlanta: 8%;
Chicago: 4;
Chicago: 10%;
Boston: 3;
Boston: 10%;
Total: N: 10;
Total: %: 9%.
Demographic: Education: Other education;
Atlanta: 2;
Atlanta: 5%;
Chicago: 3;
Chicago: 8%;
Boston: 1;
Boston: 3%;
Total: N: 6;
Total: %: 6%.
Demographic: Education: Total;
Atlanta: 40;
Atlanta: 100%;
Chicago: 39;
Chicago: 100%;
Boston: 29;
Boston: 100%;
Total: N: 108;
Total: %: 100%.
Source: NuStats, Inc.
[End of table]
During these interviews, we obtained information about the experience
of consumers with, and their opinions about, changes to their checking
accounts resulting from the check truncation process. Our interviews
included a number of standardized questions, and more tailored follow-
up questions as necessary to more fully understand their answers. All
consumers were asked about their current experience with their checking
accounts and preferred method of making retail payments. The interview
focused on consumer experience with canceled checks, substitute checks
and check images, and the possible changes to their checking accounts
since Check 21. More specifically, the structured interview of the 108
consumers included questions on the following issues: (1) bank fees
charged to them to receive canceled checks, substitute checks or image
statements; (2) instances and subsequent resolution of errors involving
their checking accounts; (3) their preferred method of receiving
information from their bank about check payments activity (such as
receiving their canceled checks, reviewing information online, or
reviewing an image statement); (4) instances in which they had to
demonstrate proof of payment using a canceled check or a check image
and their resolutions; (5) their level of concern about using a check
image as a proof of payment; and (6) whether their bank had extended
its cut-off time for accepting deposits and the consumer's opinion
about the merits of such an action. In addition, we asked nine
questions about the consumers' experience submitting complaints to
banks and federal banking regulators. This report does not contain all
the results from the consumers' interviews. We reproduced the text from
our structured interview instrument and tabulated the results from the
questions in Questions for Consumers about Check 21 Act (GAO-09-9SP).
To evaluate the benefits and costs to consumers from check truncation,
we interviewed staff from the federal banking regulators--the Board of
Governors of the Federal Reserve, the Federal Deposit Insurance
Corporation, the National Credit Union Administration, the Office of
the Comptroller of the Currency, and the Office of Thrift Supervision-
-and collected consumer complaints about the implementation of Check 21
that were submitted to these agencies from October 28, 2004, through
March 31, 2008. Our analysis of the consumer complaint data helped us
identify the issues that we pursued in our structured interviews of 108
consumers. While the regulators' consumer complaint data may be
indicative of the relative levels of different types of complaints, we
did not rely solely on these data because these voluntary reporting
systems rely on complainants to self-select themselves; therefore, the
data may not be representative of the experiences of the general
public. We also interviewed representatives from consumer advocacy
groups, including Consumers Union, the Consumer Federation of America,
and the U.S. Public Interest Research Group. Furthermore, we
interviewed officials from the American Bankers Association and third-
party processors.
The data collection instrument discussed above also included questions
about the potential benefits and costs of Check 21 for consumers. For
example, we asked the banks for information about (1) their policies on
returning canceled checks before and after Check 21; (2) the fees they
charged to consumers for the return of canceled checks and image
statements; (3) their assistance to customers in showing proof of
payment using a canceled check, a substitute check, or a check copy;
(4) the instances of expedited claims they received on substitute
checks and their resolution; and (5) the complaints they have received
about matters relating to Check 21 and whether they had changed their
cut-off times for deposits at automated teller machines or branches in
the last 2 years.
In addition, we analyzed the conclusions and the methodology applied in
the Federal Reserve Board's Report to the Congress on the Check
Clearing for the 21st Century Act of 2003, published in April 2007, to
determine whether we could use the results in our report. The study
constituted the Federal Reserve Board's assessment of the banking
industry's implementation of Check 21 to date, as well as the continued
appropriateness of the funds availability requirements of Regulation
CC.[Footnote 48] We interviewed staff from the Federal Reserve Board
about the methodology and conclusions in the report and we examined the
design, implementation, and analysis of the survey instrument used for
the study. We considered the overall strengths and weaknesses of the
Federal Reserve's data collection program, as well as specific
questionnaire items relating to Regulation CC. On the basis of our
review, we concluded that we could use the results in this report.
To determine whether consumers may incur fees for receiving canceled
checks and check images since the implementation of Check 21, we
reviewed and analyzed data purchased from Informa Research Services
(Informa) that included summary-level fee data from 2001 through
2006.[Footnote 49] The data included information on check enclosure and
imaging fees. Informa collected its data by gathering the proprietary
fee statements of banks, as well as making anonymous in-branch,
telephone, and Web site inquiries for a variety of bank fees. It also
received the information directly from its contacts at the banks. The
data are not statistically representative of the entire population of
depository institutions in the country because the company collects fee
data for particular institutions in specific geographical markets so
that these institutions can compare their fees against their
competitors. That is, surveyed institutions are self-selected into the
sample or are selected at the request of subscribers. To the extent
that institutions selected in this manner differ from those which are
not, results of the survey would not accurately reflect the industry as
a whole. Informa collects data on more than 1,500 institutions,
including a mix of banks, thrifts, credit unions, and Internet-only
banks. The institutions from which it collects data tend to be large
ones that have a large percentage of the deposits in a particular
market. Additionally, the company has access to individuals and
information from the 100 largest commercial banks.
The summary-level data Informa provided us for each data element
included the average amount, the standard deviation, the minimum and
maximum values, and the number of institutions for which data were
available to calculate the averages. They also provided these summary-
level data by institution type (banks and thrifts combined, and credit
unions) and size (as shown in table 2). In addition, Informa provided
us with data for nine specific geographic areas: California, Eastern
United States, Florida, Michigan, Midwestern United States, New York,
Southern United States, Texas, and Western United States.
Table 2: Definition of Institution Size Categories:
Institution size: Small institutions;
Asset size: Assets of less than $100 million.
Institution size: Mid-size institutions;
Asset size: Assets from $100 million to $1 billion.
Institution size: Large institutions;
Asset size: Assets of more than $1 billion.
Source: GAO analysis of Informa Research Services data.
[End of table]
We interviewed representatives from Informa to gain an understanding of
their methodology for collecting the data and the processes they had in
place to ensure the integrity of the data. Reasonableness checks were
conducted in 2007 on the data and identified any missing, erroneous, or
outlying data and Informa Research Services representatives corrected
any mistakes that were found. Also, in 2007, we compared the average
fee amounts that Informa had calculated for selected fees for 2000,
2001, and 2002 with the Federal Reserve's "Annual Report to the
Congress on Retail Fees and Services of Depository Institutions." The
averages were found to be comparable to those derived by the Federal
Reserve. While these tests did not specifically include check enclosure
and check image fees, they did confirm our assessment of the Informa
data system. Because the assessment conducted for our January 2008
report encompassed the checking fee data we used, we determined that
the Informa Research Services data were sufficiently reliable for our
current report.
We conducted this performance audit from September 2007 to October 2008
in accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objectives. We believe that
the evidence obtained provides a reasonable basis for our findings and
conclusions based on our audit objectives.
[End of section]
Appendix II: Econometric Analysis of Check 21 for Economic Efficiency
in the Federal Reserve's Check Services:
The Check Clearing for the 21st Century Act of 2003 (Check 21) was
intended to make the check payment system more efficient and less
costly by facilitating wider use of electronic check processing without
demanding that any bank change its current check collection
practices.[Footnote 50] Prior to Check 21, a bank was required to
present an original paper check to the paying bank for payment unless
the paying bank agreed to accept presentment in some other form. This
required the collecting bank to enter into agreements with all or
nearly all of the banks to which it presented checks. Because of these
impediments, banks were deterred from making the necessary electronic
check processing investments. Check 21 addressed these impediments by
authorizing a new paper negotiable instrument (a substitute check),
which is the legal equivalent of the original check. Other than
accepting the substitute check, the act does not require banks to adopt
electronic check processing, but it enables banks that want to truncate
or remove the original paper checks from the check-collection system to
do so more easily. Check 21 facilitates electronic check processing by
allowing banks to use imaging technology for collection and create
substitute checks from those images for delivery to banks that do not
accept checks electronically.
To assess the implications for economic efficiency in the Federal
Reserve System's (Federal Reserve) check processing since Check 21 took
effect in October 2004, we conducted a standard econometric analysis of
the Federal Reserve's quarterly accounting cost and volume data for the
period from 1994 through 2007. This approach allowed us to model total
check operating costs as a function of the total check presentment
volume and the timing of Check 21, while separating cost effects from
other relevant factors such as check return volume, number of check
clearing offices, and labor wages.
Description of the Econometric Models:
As suggested by microeconomic theory of the firm, we model the Federal
Reserve's total cost for its check clearing operations as a function of
outputs and input prices as shown in equation (1).[Footnote 51]
(1) 1nCt =á0 + Âbk 1n (Pk) + á1 1n (Nt ) + á2 1n (Rt) + á3 1n(0t) + á4
DC21+et:
The total check operating cost at time t (Ct) depends on the number of
checks (items) processed during that period (Nt) and the number of
return items (Rt). Total operating cost is expected to have a positive
relationship with both the total number of items processed and the
number of return items; that is, positive á1 and á2 in equation
(1).[Footnote 52]
Concurrently with the growth in electronic processing, including check
truncation, the Federal Reserve has been consolidating its check
processing operations, reducing the number of check processing sites
from 45 offices in 2000 to a planned 4 offices by early 2010. While a
reduction in the number of check processing offices (Ot) is expected to
result in savings, the Federal Reserve has reported that it incurred
consolidation and reorganization charges. Thus, the expected sign for
the coefficient of Ot [á3 in equation (1)] is ambiguous; it may be
positive in the case of a cost savings or negative in the case of an
increase in total costs.
The coefficient of primary interest is that of Dc21, (á4). The dummy
variable Dc21 is constructed to have a value of 1 for periods on and
after October 2004 when Check 21 took effect and a value of zero
otherwise. As Check 21 was intended to facilitate electronic clearing
of checks, the hypothesis is that after the effective date of Check 21
the Federal Reserve's total operating costs would decrease; that is, a
negative á4 in the estimation. Consistent with microeconomic theory, we
expect an increase in input prices (pk) will lead to an increase in
total cost. For example, higher labor wage rates are expected to lead
to higher total cost, seen as positive coefficients for input prices in
the estimation.
Based on econometric studies, including some that specifically
considered economies of scale for check processing, we modified the
basic approach of equation (1) to control for quarterly fluctuations
and trends over time, and to consider the potential effects of Check 21
on the presence of scale economies in check clearing
operations.[Footnote 53] That is, we modified our equation (1) to
include a structural break associated with Check 21.[Footnote 54] This
specification better reflects the cost effect of the technology shift
from processing paper checks to processing checks electronically. A
negative coefficient of the output variable multiplied by the dummy
variable for Check 21 will imply a more cost-efficient cost structure
for the period after Check 21; a positive coefficient will imply a
higher cost structure. QTR, time and time2 are dummy variables to
control for quarterly fluctuation and trends over time.
(2) lnCt = á0 + Sbk 1n (Pk) + á1 1n (Nt ) + á2 1n(Rt) + á3 1n(0t) + á4
DC21l1time + l2time2:
3:
+Sjj QTR j + [ápo + SbPk 1n (Pk) + áp1 1n (Nt ) + áp21n(Rt) + áp3 1n(0t
) + áp4DC21]x Dc21 +et:
1:
Results:
We estimated equation (2) with quarterly data from the Federal
Reserve's Planning and Control System (PACS) for the period from 1994
through 2007. Table 3 shows the summary statistics for selected
variables.
Table 3: Summary Statistics of Some Selected Variables:
Variables: Logarithm of total check operating cost;
Mean: 18.94;
Maximum: 19.23;
Minimum: 18.67;
Standard deviation: 0.16.
Variables: Logarithm of total presentment[A];
Mean: 22.05;
Maximum: 22.21;
Minimum: 21.61;
Standard deviation: 0.17.
Variables: Logarithm of check clearing offices;
Mean: 3.66;
Maximum: 3.85;
Minimum: 2.94;
Standard deviation: 0.28.
Variables: Logarithm of returned checks;
Mean: 17.50;
Maximum: 17.69;
Minimum: 17.14;
Standard deviation: 0.16.
Variables: Logarithm of price deflator for equipment and software[B];
Mean: 4.62;
Maximum: 4.76;
Minimum: 4.55;
Standard deviation: 0.08.
Variables: Logarithm of wage[C];
Mean: 2.61;
Maximum: 2.83;
Minimum: 2.39;
Standard deviation: 0.13.
Variables: Logarithm of Gross Domestic Product[D];
Mean: 4.63;
Maximum: 4.79;
Minimum: 4.50;
Standard deviation: 0.09.
Source: GAO analysis of PACS data.
[A] Total presentment includes presentment by traditional paper checks,
legacy paper checks, check images, and substitute checks.
[B] The price deflator for equipment and software by nonresidential
producers (2000=100).
[C] Average hourly earning for all the private sector in 2000 dollars.
[D] Gross Domestic Product deflator (2000=100).
[End of table]
Results Using Total Check Presentment:
We estimated the logarithm of total check processing cost against the
logarithms of total presentment items--image, paper, legacy, and
substitute--and other related variables.[Footnote 55] Table 4 presents
the results. The basic specification in table 4, which does not account
for a possible different cost structure in check processing, yields
mostly statistically insignificant coefficients. However, the
coefficient for the total number of items presented is significant and
positive, implying that a 1 percent increase in total presentment will
result in a 1.34 percent increase in total cost.
However, the coefficient for the Check 21 dummy variable (Check21),
while negative, is not statistically significant. This result does not
provide any support for the hypothesis that the introduction of Check
21 led to a decrease in Federal Reserve costs, although it is not
possible to determine the extent to which this may be driven by the
concurrent consolidation of Federal Reserve check services sites.
Table 4: Estimation of Logarithm of Total Check Operating Cost, 1994-
2007:
Description: Constant;
Basic: Coefficient: -23.41;
Basic: Std. error: 21.16;
With structural break: Coefficient: -17.73;
With structural break: Std. error: 23.39.
Description: Logarithm of total presentment;
Basic: Coefficient: 1.34;
Basic: Std. error: 0.43;
With structural break: Coefficient: 0.84;
With structural break: Std. error: 0.58.
Description: Logarithm of total presentment*Check21;
Basic: Coefficient: Basic: Std. error: With structural break:
Coefficient: -0.25;
With structural break: Std. error: 1.21.
Description: Logarithm of returned checks;
Basic: Coefficient: 0.35;
Basic: Std. error: 0.19;
With structural break: Coefficient: 0.49;
With structural break: Std. error: 0.28.
Description: Logarithm of returned checks*Check 21;
Basic: Coefficient: Basic: Std. error: With structural break:
Coefficient: -0.95;
With structural break: Std. error: 0.90.
Description: Logarithm of number of offices;
Basic: Coefficient: -0.098;
Basic: Std. error: 0.17;
With structural break: Coefficient: -0.041;
With structural break: Std. error: 0.32.
Description: Logarithm of number of offices*Check 21;
Basic: Coefficient: Basic: Std. error: With structural break:
Coefficient: -0.13;
With structural break: Std. error: 0.48.
Description: Logarithm of wage;
Basic: Coefficient: 1.01;
Basic: Std. error: 1.14;
With structural break: Coefficient: 3.77;
With structural break: Std. error: 1.96.
Description: Logarithm of wage*Check 21;
Basic: Coefficient: Basic: Std. error: --;
With structural break: Coefficient: -5.42;
With structural break: Std. error: 4.37.
Description: Logarithm of price deflator for equipment and software;
Basic: Coefficient: 0.025;
Basic: Std. error: 1.70;
With structural break: Coefficient: 0.63;
With structural break: Std. error: 2.16.
Description: Logarithm of price deflator for equipment and
software*Check 21;
Basic: Coefficient: Basic: Std. error: --;
With structural break: Coefficient: -3.65;
With structural break: Std. error: 9.14.
Description: Logarithm of Gross Domestic Product;
Basic: Coefficient: 0.99;
Basic: Std. error: 3.99;
With structural break: Coefficient: 0.55;
With structural break: Std. error: 4.48.
Description: Logarithm of Gross Domestic Product *Check21;
Basic: Coefficient: -- ;
Basic: Std. error: --;
With structural break: Coefficient: 2.27;
With structural break: Std. error: 4.25.
Description: Check 21[A];
Basic: Coefficient: -0.023;
Basic: Std. error: 0.048;
With structural break: Coefficient: 43.17;
With structural break: Std. error: 46.36.
Description: Quarter I dummy;
Basic: Coefficient: -0.008;
Basic: Std. error: 0.036;
With structural break: Coefficient:
-0.052;
With structural break: Std. error: 0.043.
Description: Quarter 2 dummy;
Basic: Coefficient: -0.036;
Basic: Std. error: 0.019;
With structural break: Coefficient:
-0.039;
With structural break: Std. error: 0.022.
Description: Quarter 3 dummy;
Basic: Coefficient: -0.0023;
Basic: Std. error: 0.016;
With structural break: Coefficient:
-0.006;
With structural break: Std. error: 0.018.
Description: Trend;
Basic: Coefficient: -0.017;
Basic: Std. error: 0.028;
With structural break: Coefficient: -0.031;
With structural break: Std. error: 0.030.
Description: Trend squared;
Basic: Coefficient: 0.00039;
Basic: Std. error: 0.00016;
With structural break: Coefficient: 0.0004;
With structural break: Std. error: 0.00017.
Description: AR(1)[B];
Basic: Coefficient: 0.41;
Basic: Std. error: 0.16;
With structural break: Coefficient: 0.36;
With structural break: Std. error: 0.20.
Description: Number of observations;
Basic: Coefficient: 55;
Basic: Std. error: [Empty];
With structural break: Coefficient: 55;
With structural break: Std. error: [Empty].
Description: R-squared;
Basic: Coefficient: 0.95;
Basic: Std. error: [Empty];
With structural break: Coefficient: 0.96;
With structural break: Std. error: [Empty].
Description: Adj R-squared;
Basic: Coefficient: 0.94;
Basic: Std. error: [Empty];
With structural break: Coefficient: 0.94;
With structural break: Std. error: [Empty].
Description: Log likelihood;
Basic: Coefficient: 107.43;
Basic: Std. error: [Empty];
With structural break: Coefficient: 111.75;
With structural break: Std. error: [Empty].
Description: F-statistic;
Basic: Coefficient: 61.83;
Basic: Std. error: [Empty];
With structural break: Coefficient: 42.57;
With structural break: Std. error: [Empty].
Description: D-W stat;
Basic: Coefficient: 1.68;
Basic: Std. error: [Empty];
With structural break: Coefficient: 1.71;
With structural break: Std. error: [Empty].
Source: GAO analysis of PACS data.
[A] One on and after October 2004; zero otherwise.
[B] AR (1) is the coefficient of the first order of autocorrelation,
which is to control of the correlation between the current and the
previous periods.
[End of table]
Table 4 also shows the results of the estimation incorporating a
structural break in the cost function for periods before and after the
act as described in equation (2). Though insignificant, the coefficient
of total presentment is positive and less than 1, and the coefficient
of the interacted variable of total presentment and Check 21 dummy is
negative (-0.25). If significant, the sum of two coefficients would
imply that the cost structure for the check operation in the post-Check
21 period would be different from the pre-Check 21 period. However, the
relatively short time series data for the post-Check 21 period increase
the standard errors for all the coefficients of the interacted
variables. Also, although insignificant, the coefficient of the Check
21 dummy is positive, implying that the total cost, on average, is
lower in periods before Check 21 than after.
In addition to the estimation results shown in table 4, we estimated
alternative functional forms used in other similar studies for the
relationships in equation (2).[Footnote 56] Because these functional
forms generally require constructing a substantial number of interacted
variables, the subsequent multicollinearity and the limited data
available make the results subject to high estimation errors and thus
difficult from which to draw clear inferences. We also tested the
effects on the estimates of imposing a constraint suggested by economic
theory.[Footnote 57] The standard errors for most of the coefficient
estimates decrease, suggesting a decrease in multicollinearity, but the
results are otherwise similar to the results without the constraint in
table 4.
Potential Limitations of the Analysis:
While we believe that this analysis provides a reasonable basis for our
findings and conclusions regarding the effects of Check 21 on Federal
Reserve costs, we recognize the limitations inherent in the analysis.
First, our econometric analysis only uses Federal Reserve data on check
operations, which may not be representative of the operations of
private financial institutions. Second, the time series data on post-
Check 21 cost and volume may not be long enough to reliably estimate
the effect on cost of the accelerating use of electronic presentment
and clearing, particularly given the changes in technology embodied in
electronic presentment and check truncation. 's check processing.
Also, as previously mentioned, the Federal Reserve's ongoing effort to
close check clearing office facilities has resulted in one-time
consolidation and reorganization charges. These charges are included in
the total cost operating costs, and although we try to control for
their effect by including the number of offices variable, it is
plausible that the positive sign of the Check 21 dummy in our
estimations may be a result of these charges included in the total
costs. Similarly, our analysis implicitly assumes that the Federal
Reserve's consolidation decisions are independent of the volume of
checks that it processes. However, the data are not sufficient to
explicitly model a relationship between the volume of checks and
expectations about future volumes.
[End of section]
Appendix III Comments from the Board of Governors of the Federal
Reserve System:
Board Of Governors Of The Federal Reserve System:
Washington, D.C. 20551:
Louise L. Roseman:
Director:
Division Of Reserve Bank Operations And Payment Systems:
Email: Louise.Roseman@frb.gov:
Phone: (202) 452-2789:
Fax: (202) 452-2746:
October 16, 2008:
Ms. Yvonne D. Jones:
Director:
Financial Markets and Community Investment:
United States Government Accountability Office:
441 G Street, NW:
Washington, DC 20548:
Dear Ms. Jones:
We appreciate the opportunity to comment on the GAO's report titled
Check 21 Act: Most Consumers Have Accepted and Banks Are Progressing
Towards Full Adoption of Check Truncation. We agree with the GAO's
overall conclusion that, over the past four years, the banking industry
has made substantial progress towards establishing an end-to-end
electronic check-processing environment. Today, more than three-
quarters of checks deposited with the Federal Reserve Banks for
collection are deposited electronically, and more than half are
presented electronically. The Federal Reserve Banks expect that by year-
end 2009, more than 90 percent of their check deposits and presentments
will be electronic.
This ongoing transformation to an end-to-end electronic check-
processing environment has not been without cost. The banking industry
and the Federal Reserve Banks have made significant technological
investments to facilitate an electronic check-clearing system and have
incurred incremental transition costs associated with processing both
paper and electronic checks. The Federal Reserve Banks' investments,
however, have enabled them to significantly reduce their transportation
costs and paper check-processing infrastructure. These cost reductions
have been critical to the Reserve Banks' ability to recover all of
their actual and imputed costs of providing check services from 2005
through 2007 and earn a net income of $326 million.
We are pleased that consumers are beginning to obtain the benefits
associated with Check 21 and that this major transformation from paper
to electronic check clearing is being accomplished with few consumer
complaints. The report reviews a number of these consumer benefits:
faster funds availability on check deposits due to later deposit
deadlines, quicker access to account information, and improved customer
service. In addition, as the report notes, another important consumer
benefit that has flowed from Check 21 has been a shorter maximum
permissible hold period on an increasing number of check deposits. By
facilitating the electronic collection of checks, Check 21 has enabled
the Reserve Banks to close most of their check-processing offices,
resulting in the consolidation of many Federal Reserve check-
processing regions. Because Congress tied the determination of whether
a check is local or nonlocal to whether the depositary bank and the
paying bank are located in the same check-processing region, many
checks that were previously classified as nonlocal checks subject to a
five-day maximum permissible hold are now classified as local checks
subject to a maximum two-day hold period. It is likely that within the
next several years, all checks will be classified as local, subject to
the shorter permissible hold period.
Again, we appreciate the opportunity to review and comment on the GAO's
report and the efforts and professionalism of the GAO's team in
conducting this study.
Sincerely,
Signed by:
Louise L. Roseman:
[End of section]
Appendix IV: GAO Contact and Staff Acknowledgments:
GAO Contact:
Yvonne D. Jones (202) 512-8678 or jonesy@gao.gov:
Staff Acknowledgments:
The following individuals made key contribution to this report: Debra
R. Johnson, Assistant Director; Joanna Chan; Philip Curtin; Nancy
Eibeck; Terence Lam; James McDermott; Carl Ramirez; Barbara Roesmann;
and Paul Thompson.
[End of section]
Footnotes:
[1] The Federal Reserve System, 2007 Federal Reserve Payments Study:
Noncash Payment Trends in the United States 2003-2006 (Washington,
D.C.: Dec. 10, 2007), 6. The study is part of an ongoing effort by the
Federal Reserve System to measure trends in noncash payment in the
United States.
[2] Pub. L. No. 108-100 (Oct. 28, 2003). For this report, we refer to
commercial banks, thrifts, and credit unions collectively as banks.
[3] A check image is an electronic or digital image of an original
check that is created by a depositor, a bank, or other participant in
the check collection process.
[4] For this report, we define economic efficiency as an economically
efficient production that is organized to minimize the ratio of inputs
to outputs. Production is economically efficient when goods are
produced at minimum cost in money and resources. This typically occurs
where input prices are used to find the least-expensive process.
[5] Interbank checks are those in which the bank of first deposit and
the paying bank are different. On-us checks are deposited or cashed at
the same bank on which they are drawn.
[6] See U.C.C. §§ 4-301, 302. The Uniform Commercial Code is a set of
model laws adopted and enacted by the states that govern commercial and
financial activities.
[7] See 12 C.F.R. § 229.12. Local checks are checks in which the bank
of first deposit and the paying bank are located in the same Federal
Reserve check-processing region. Nonlocal checks are checks in which
the bank of first deposit and the paying bank are located in different
Federal Reserve check-processing regions.
[8] 12 C.F.R. § 229.2(ddd).
[9] GAO, Retail Payments Issues: Experience with Electronic Check
Presentment, GAO/GGD-98-145 (Washington, D.C.: July 14, 1998).
[10] GAO/GGD-98-145.
[11] Geoffrey R. Gerdes and Jack K. Walton II, "The Use of Checks and
Other Noncash Payment Instruments in the United States," Federal
Reserve Bulletin (August 2002), 360.
[12] In a check conversion, the paper check is used only as an
information source and is converted into an ACH payment so that it can
be processed on the ACH network. The ACH is an electronic batch
processing system by which payment orders are exchanged among banks. By
agreement, consumer checks can be converted into ACH payments by
merchants at the point of sale or by billers that receive check
remittances. Checks that are converted into ACH payments never enter
the check collection system. The Federal Reserve reported that 2.6
billion paper checks were converted into ACH transactions in 2006.
[13] EBT was devised in the 1980s to meet the needs of the U.S.
Department of Agriculture's Food Stamp Program. Its initial purpose was
to transfer federal benefits electronically to eligible recipients
under certain entitlement and grant programs.
[14] FedForward is a service in which checks are deposited
electronically for collection using image cash letters. (Cash letter is
a group of checks packaged and sent by a bank to another bank,
clearinghouse, or a Federal Reserve office. A cash letter is
accompanied by a list containing the dollar amount of each check, the
total amount of the checks, and the number of checks sent with the cash
letter.) FedReceipt is a service in which the paying bank agrees to the
electronic presentment of checks with accompanying images. FedReturn is
a service in which checks to be returned are sent in image cash letters
to the Federal Reserve, which will return the checks either in an
electronic file or as substitute checks to the banks of first deposit.
[15] Under the Monetary Control Act of 1980, Pub. L. No. 96-221 (March
31, 1980) (12 U.S.C. 248a) the Federal Reserve is to set fees charged
to banks for providing priced services including check services to
recover, over the long run, all direct and indirect costs of providing
the services plus imputed costs, including the interest on items
credited before actual collection (float), and a private-sector
adjustment factor set to reflect costs that a private-sector firm would
face. See 12 U.S.C. § 248a.
[16] Board of Governors of the Federal Reserve System, 2008 Annual
Report: Budget Review (Washington, D.C.: May 2008), 16.
[17] Board of Governors of the Federal Reserve System, 2007 Annual
Report: Budget Review (Washington, D.C.: April 2007), 17-18.
[18] Board of Governors of the Federal Reserve System, 2006 Annual
Report: Budget Review (Washington, D.C.: April 2006), 22.
[19] See for example, Ernst R. Berndt, The Practice of Econometrics:
Classic and Contemporary (Mass: Addison-Wesley Publishing, 1996),
chapter 3; Robert M. Adams, Paul W. Bauer, and Robin C. Sickles,
Federal Reserve Bank of Cleveland, "Scope and Scale Economies in
Federal Reserve Payment Processing," Working Paper 02-13 (November
2002); David B. Humphrey, "Scale Economies at Automated Clearing
House," Journal of Bank Research (Summer 1981), 71-81; and Paul W.
Bauer and Dianna Hancock, "Scale Economies and Technological Change in
the Federal Reserve ACH Payment Processing," Federal Reserve Bank of
Cleveland Economic Review (1995) vol. 31, no. 3, 14-29.
[20] However, the magnitude of the cost effects that are based on this
analysis vary considerably with different functional forms and are
particularly sensitive to the explanatory variables included in the
estimation process. Further, many of the explanatory variables are
highly correlated, which inherently limits efforts to differentiate
their effects on total costs. This issue, to which the econometrics
literature refers as mulitcollinearity, also implies that the results
could change if data on subsequent quarters were added.
[21] The Retail Payments Office, located at the Federal Reserve Bank of
Atlanta, directs the Federal Reserve's Banks' retail payment
activities.
[22] Pub. L. 100-86, title VI, Aug. 10, 1987, 101 Stat. 635 (12 U.S.C.
4001 et. seq.) The funds availability schedules are set forth at 12
U.S.C. 4002. The Federal Reserve's implementing regulations are set
forth at 12 C.F.R. Part 229 (Regulation CC), Subpart B.
[23] In this instance, float refers to the interest associated with the
quicker collection of funds associated with check deposits.
[24] See Subcommittee on Financial Institutions and Consumer Credit,
House Committee on Financial Services, Testimony of Louise L. Roseman,
Director, Division of Reserve Bank Operations and Payment Systems, 109
Congr. 1st sess., April 20, 2005.
[25] ABA Deposit Account Fraud Survey Report 2007 Edition. The survey
was conducted from February through July of 2007. According to the
report, a total of 176 institutions returned completed survey forms and
most of the information presented in the report was for calendar year
2006.
[26] GAO/GGD-98-145.
[27] For this specific question, one interview participant did not
respond to the question. Thus, the total number of responses for this
question was 107, not 108.
[28] Furthermore, 69 of the 108 consumers, or about 64 percent, stated
that they conducted some type of online banking, including activities
such as account balancing, bill payment, and funds transfers.
[29] Double-posting or duplicate payment of a check refers to the same
check being presented to the paying bank twice.
[30] We used data from the Federal Deposit Insurance Corporation, the
Office of the Comptroller of the Currency, the Office of Thrift
Supervision, and the Federal Reserve Board. We were unable to obtain
specific consumer complaint data related to Check 21 from the National
Credit Union Administration.
[31] See 12 C.F.R. 229.19(a)(5)(ii).
[32] Board of Governors of the Federal Reserve System, Report to the
Congress on the Check Clearing for the 21st Century Act of 2003 (April
2007), 17.
[33] Check Clearing for the 21st Century, 15.
[34] Check Clearing for the 21st Century, 13.
[35] Check Clearing for the 21st Century, 13.
[36] See, e.g., Uniform Commercial Code Articles 4-401 and 4-402.
[37] Additional elements also must be satisfied. See 12 U.S.C. 5006(a)
see also 12 C.F.R. § 229.54. Warranties pertaining to the substitute
check provide that the substitute check meets all the requirements for
legal equivalence and that there is no duplicate payment of a check.
See 12 U.S.C. 5004; see also, Regulation CC, 12 C.F.R. §§
229.52.229.53.
[38] If the bank has not acted on the claim before the end of the tenth
business day after the banking day on which the bank received the
claim, it must provisionally recredit the consumer's account up to the
lesser of the amount of the substitute check or $2,500. The bank must
recredit any remaining balance greater than $2,500 no later than the
forty-fifth calendar day after the banking day on which the bank
received the claim. 12 U.S.C. § 5006.
[39] We note that the bank with less than 1,000 claims is among the
largest in the country, handling more than 3.2 billion checks in 2007.
In addition, the institution reported that 75 percent of the claims
were resolved in the customer's favor.
[40] See GAO, Bank Fees: Federal Banking Regulators Could Better Ensure
that Consumers Have Required Disclosure Documents Prior to Opening
Checking or Savings Accounts, GAO-08-281 (Washington, D.C.: Jan. 31,
2008). The Informa data typically were gathered from retail banks with
large market shares in specific areas and are not statistically
generalizable to other institutions.
[41] Informa defined the check enclosure fee as the fee charged to have
cancelled checks returned with the statement and the check image fee as
the fee charged to have images of cancelled checks returned with the
statement.
[42] Informa defined a non-interest, free checking account as a free
checking account that does not earn interest, has no monthly or
transaction fees (such as debit per check fees or ATM/check card debit
fees), and has no balance requirements.
[43] Informa defined a senior checking account as a checking account
exclusively for seniors (persons age 65 or older) that can be interest
or non-interest bearing.
[44] See GAO-08-281.
[45] See GAO-08-281. See also Eric Halperin, et al., Center for
Responsible Lending, Debit Card Danger: Banks Offer Little Warning and
Few Choices as Customers Pay a High Price for Debit Card Overdrafts
(Jan. 25, 2007).
[46] For this report, commercial banks, financial institutions,
thrifts, and credit unions will be collectively referred to as banks.
[47] Corporate credit unions are nonprofit financial cooperatives that
are owned by natural person credit unions (that is, credit unions whose
members are individuals), and provide lending, investment, and other
financial services to these credit unions. Additionally, corporate
credit unions offer automated settlement, securities safekeeping, data
processing, accounting, and electronic payments services, which are
similar to the correspondent services that large commercial banks have
traditionally provided to smaller banks. Bankers' banks were created as
unique entities to provide community banks a noncompeting institution
to offer correspondent and other financial services.
[48] Regulation CC implements both the Expedited Funds Availability Act
of 1987 and the Check Clearing for the 21st Century Act of 2003. Among
other things, Regulation CC sets forth requirements that banks make
funds deposited in transaction accounts like checking accounts
according to specified time schedules and banks disclose their funds
availability policies to their customers. The regulation also
establishes rules for the collection and the return of unpaid checks.
Subpart D of the Regulation CC describes the requirements for
substitute checks, the reconverting bank's duties, and expedited re-
credit for consumers and banks, among other things.
[49] We originally purchased these data in connection with our analysis
of bank fees. For more information, see GAO, Bank Fees: Federal Banking
Regulators Could Better Ensure That Consumers Have Required Disclosure
Documents Prior to Opening Checking or Savings Accounts, GAO-08-281
(Washington, D.C.: Jan. 31, 2008).
[50] For this report, we refer to banks, thrifts, and credit unions
collectively as banks.
[51] Many microeconomic textbooks have detailed discussions on cost
function. For example, see Hal R. Varian, Microeconomic Analysis, 3rd
edition (New York, N.Y.: W.W. Norton & Company, 1993), chapter 5.
[52] We do not account for float in the model. Float refers to the time
between the payment transaction and the debiting of funds from an
account.
[53] Some of these studies include Ernst R. Berndt, The Practice of
Econometrics: Classic and Contemporary (Mass: Addison-Wesley
Publishing, 1996), chapter 3; Robert M. Adams, Paul W. Bauer, and Robin
C. Sickles, Federal Reserve Bank of Cleveland, "Scope and Scale
Economies in Federal Reserve Payment Processing," Working Paper 02-13
(November 2002); David B. Humphrey, "Scale Economies at Automated
Clearing House," Journal of Bank Research (Summer 1981), 71-81; and
Paul W. Bauer and Dianna Hancock, "Scale Economies and Technological
Change in the Federal Reserve ACH Payment Processing," Federal Reserve
Bank of Cleveland Economic Review (1995) vol. 31, no. 3, 14-29.
[54] Because the focus of our analyses is to examine the relationship
between Check 21and total cost of check processing, we used a
parsimonious specification of a simple log-linear cost function
(equation 2) in our estimation instead of more sophisticated
specifications in other studies.
[55] Because deposit and presentment are almost mirror images of each
other, we opt to use presentment as the independent variables in our
estimations.
[56] For example, we estimated some parsimonious versions of the
separable quadratic cost function as used in the study by Robert M.
Adams, Paul W. Bauer, and Robin C. Sickles, Federal Reserve Bank of
Cleveland, "Scope and Scale Economies in Federal Reserve Payment
Processing," Working Paper 02-13 (November 2002).
[57] This constraint of linear homogeneity is to ensure that the cost
function is consistent with microeconomic theory--if one doubles the
price of the input factors, one would expect to double total costs.
Mathematically, it means that the sum of bk should add up to 1,
k=N
 bk = 1
k=1
To impose this constraint, we made some adjustments to the total costs
and input price. See William H. Green, Econometric Analysis (Prentice
Hall, N.J.: 1993), 503-507.
GAO's Mission:
The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people.
GAO examines the use of public funds; evaluates federal programs and
policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding
decisions. GAO's commitment to good government is reflected in its core
values of accountability, integrity, and reliability.
Obtaining Copies of GAO Reports and Testimony:
The fastest and easiest way to obtain copies of GAO documents at no
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each
weekday, GAO posts newly released reports, testimony, and
correspondence on its Web site. To have GAO e-mail you a list of newly
posted products every afternoon, go to [hyperlink, http://www.gao.gov]
and select "E-mail Updates."
Order by Mail or Phone:
The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or
more copies mailed to a single address are discounted 25 percent.
Orders should be sent to:
U.S. Government Accountability Office:
441 G Street NW, Room LM:
Washington, D.C. 20548:
To order by Phone:
Voice: (202) 512-6000:
TDD: (202) 512-2537:
Fax: (202) 512-6061:
To Report Fraud, Waste, and Abuse in Federal Programs:
Contact:
Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]:
E-mail: fraudnet@gao.gov:
Automated answering system: (800) 424-5454 or (202) 512-7470:
Congressional Relations:
Ralph Dawn, Managing Director, dawnr@gao.gov:
(202) 512-4400:
U.S. Government Accountability Office:
441 G Street NW, Room 7125:
Washington, D.C. 20548:
Public Affairs:
Chuck Young, Managing Director, youngc1@gao.gov:
(202) 512-4800:
U.S. Government Accountability Office:
441 G Street NW, Room 7149:
Washington, D.C. 20548: