Full text of Federal Reserve Payments Study : Retail Payments Research Project : A Snapshot of the U.S. Payments Landscape

View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

RETAIL PAYMENTS RESEARCH PROJECT

A Snapshot of the U.S. Payments Landscape

Depository Financial Institution Check Study

Check Sample Study

Electronic Payment Instruments Study

Research Sponsored by the Federal Reserve System.

2
Federal Reserve Project Team for the Retail Payments Research Project:
Hank Bourgaux
Senior Vice President
Financial Services
Federal Reserve Bank of St. Louis
Geoffrey R. Gerdes
Economist
Board of Governors of the Federal Reserve System
Richard Oliver
Senior Vice President and Retail Product Manager
Federal Reserve Bank of Atlanta
Kathy O. Paese
Assistant Vice President
Financial Services
Federal Reserve Bank of St. Louis
Darrel W. Parke
Chief
Micro Statistics Section
Division of Research and Statistics
Board of Governors of the Federal Reserve System
Terrence J. Roth
Vice President
Check Services
Federal Reserve Bank of Cleveland
Samuel M. Slowinski
Statistician
Micro Statistics Section
Division of Research and Statistics
Board of Governors of the Federal Reserve System
Jack K. Walton II
Assistant Director
Division of Reserve Bank Operations and Payment Systems
Board of Governors of the Federal Reserve System
Stuart E. Weiner
Vice President & Economist
Payments System Research
Federal Reserve Bank of Kansas City

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

3
Depository Financial Institution Check Study and Check Sample Study Conducted by
Global Concepts and Westat
Electronic Payment Instruments Study Conducted by
Dove Consulting

Report Authors
Global Concepts
Stephen K. Ledford, Senior Principal
David C. Stewart, Principal
Tom Murphy, Principal
Timothy Arscott-Mills, Director
Anne Battle, Senior Research Writer/Editor

Westat
W. Wade Martin, Project Director
G. Hussain Choudhry, Senior Statistician
Pat Dean Brick, Survey Methodologist
Howard King, Operations Manager

Dove Consulting
Ed Bachelder, Director of Research and Analytics
Beth Costa, Director Financial Services
Mark Haddad, Consultant
Jessica Ip, Research Analyst

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

4
Table of Contents
1

EXECUTIVE SUMMARY................................................................................................................................9
1.1
1.2
1.2.1
1.2.2
1.3
1.3.1
1.4
1.4.1
1.4.2
1.4.3
1.4.4
1.4.5
1.5
1.6

INTRODUCTION............................................................................................................................................23
2.1

INTRODUCTION ............................................................................................................................................9
SIGNIFICANT RESULTS: THE DEPOSITORY FINANCIAL INSTITUTION CHECK STUDY..................................10
Check Volume and Value by Clearing Methods ..................................................................................11
Returned Check Volume and Value .....................................................................................................12
SIGNIFICANT RESULTS: THE CHECK SAMPLE STUDY ................................................................................12
Categorization of Check Payments by Purpose ...................................................................................14
SIGNIFICANT FINDINGS: ELECTRONIC PAYMENT INSTRUMENTS STUDY ....................................................17
General Purpose and Private Label Credit Card ................................................................................19
Online and Offline Debit Cards...........................................................................................................19
ACH .....................................................................................................................................................19
EBT ......................................................................................................................................................20
Emerging Payments .............................................................................................................................20
NON-CASH PAYMENTS MARKET THEN VS. NOW ......................................................................................20
IMPLICATIONS AND OBSERVATIONS ..........................................................................................................21
SURVEY METHODOLOGIES ........................................................................................................................23

THE DEPOSITORY FINANCIAL INSTITUTION CHECK STUDY.......................................................25
3.1
METHODOLOGY.........................................................................................................................................25
3.1.1
Sample Design .....................................................................................................................................25
3.1.1.1
3.1.1.2
3.1.1.3

3.1.2
3.1.3
3.1.4

Sampling Frame of the Financial Institutions ............................................................................................ 26
Certainty Strata .......................................................................................................................................... 26
Sample Size and Sample Allocation .......................................................................................................... 26

Sample Weighting ................................................................................................................................27
Estimation ............................................................................................................................................27
Designing the Survey Instrument .........................................................................................................27

3.1.4.1
3.1.4.2
3.1.4.3
3.1.4.4

Paid Checks ............................................................................................................................................... 28
Outgoing Returns ....................................................................................................................................... 28
Routing Transit Numbers (RTs)................................................................................................................. 28
Pretesting the Survey Instrument ............................................................................................................... 28

3.1.5
The Survey Instrument .........................................................................................................................28
3.1.6
Reporting Period: March 1, 2001 - April 30, 2001 .............................................................................29
3.1.7
Data Collection....................................................................................................................................29
3.2
RESULTS AND ANALYSIS ...........................................................................................................................30
3.2.1
Survey Response ..................................................................................................................................30
3.2.2
Aggregate Paid Check Volume and Value...........................................................................................30
3.2.3
Check Volume and Value by Clearing Method ....................................................................................31
3.2.4
Outgoing Returns Volume and Value Data..........................................................................................32
4

THE CHECK SAMPLE STUDY (CSS).........................................................................................................33
4.1
METHODOLOGY.........................................................................................................................................33
4.1.1
CSS Sample Design..............................................................................................................................33
4.1.1.1

4.1.2
4.1.3
4.1.4

4.1.4.1
4.1.4.2
4.1.4.3

4.1.5

The Sampling Parameters Request Form ................................................................................................... 35
Retrieval of Randomly Selected Checks .................................................................................................... 36
Photo Retrieval Latitude ............................................................................................................................ 36

Designing the Survey Instrument .........................................................................................................37

4.1.5.1

4.1.6

Sample Size and Sample Allocation .......................................................................................................... 34

CSS Sample Weighting.........................................................................................................................35
Estimation ............................................................................................................................................35
Random Sampling in the Field.............................................................................................................35

Pretesting the Survey Instrument ............................................................................................................... 37

The Survey Instrument .........................................................................................................................37

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

5
4.1.7
4.1.8

Reporting Period: May 1, 2000 - April 30, 2001.................................................................................37
Check Payments Categorization ..........................................................................................................38

4.1.8.1
Payer and Payee Categories ....................................................................................................................... 38
4.1.8.2
Purpose Categories .................................................................................................................................... 39
4.1.8.3
Groupings of Categories for Data Analysis ............................................................................................... 41
4.1.8.3.1
Payer and Payee Groups ....................................................................................................................... 41
4.1.8.3.2
Purpose Groups..................................................................................................................................... 42

4.1.9

The Check Categorization Model ........................................................................................................42

4.1.9.1
4.1.9.2
4.1.9.3

Categorization of the Payer........................................................................................................................ 42
Categorization of the Payee ....................................................................................................................... 43
Categorization of the Purpose .................................................................................................................... 43

4.1.10
Data Review ....................................................................................................................................44
4.2
RESULTS AND ANALYSIS ...........................................................................................................................44
4.2.1
Survey Response ..................................................................................................................................44
4.2.2
Survey Results......................................................................................................................................45
4.2.2.1
4.2.2.2
4.2.2.3
4.2.2.4

Distribution of Check Volume ................................................................................................................... 45
Distribution of Check Value ...................................................................................................................... 47
Average Dollar Value by Payment Purpose and Counterparty .................................................................. 48
Distribution of Checks by Dollar Value Category ..................................................................................... 49

ELECTRONIC PAYMENT INSTRUMENTS STUDY ...............................................................................50
5.1
INTRODUCTION ..........................................................................................................................................50
5.1.1
Objectives ............................................................................................................................................50
5.1.2
Scope....................................................................................................................................................50
5.1.2.1

Excluded from Scope................................................................................................................................. 51

5.2
RESEARCH METHODOLOGY.......................................................................................................................52
5.2.1
Participation Rates ..............................................................................................................................52
5.3
SUMMARY OF FINDINGS ............................................................................................................................52
5.3.1
General-Purpose and Private-Label Credit Cards..............................................................................53
5.3.2
Online and Offline Debit Cards...........................................................................................................54
5.3.3
ACH .....................................................................................................................................................54
5.3.4
EBT ......................................................................................................................................................54
5.3.5
Emerging Payments .............................................................................................................................54
5.4
ANALYSIS OF FINDINGS AND METHODOLOGY ...........................................................................................54
5.4.1
General-Purpose Credit Cards............................................................................................................54
5.4.1.1
5.4.1.2
5.4.1.3

5.4.2

Private-Label Credit Cards .................................................................................................................58

5.4.2.1
5.4.2.2
5.4.2.3

5.4.3

Background................................................................................................................................................ 54
Organizations ............................................................................................................................................. 56
Survey Data: General-Purpose Credit Cards.............................................................................................. 56
Background................................................................................................................................................ 58
Organizations ............................................................................................................................................. 58
Survey Data: Private-Label Credit Cards................................................................................................... 58

Debit Cards..........................................................................................................................................60

5.4.3.1
Background................................................................................................................................................ 60
5.4.3.2
Online Debit............................................................................................................................................... 61
5.4.3.2.1
Organizations........................................................................................................................................ 61
5.4.3.2.2
Survey Data: Online Debit.................................................................................................................... 61
5.4.3.3
Offline Debit .............................................................................................................................................. 62
5.4.3.3.1
Organizations........................................................................................................................................ 62
5.4.3.3.2
Survey Data: Offline Debit ................................................................................................................... 62

5.4.4

Automated Clearinghouse....................................................................................................................63

5.4.4.1
5.4.4.2
5.4.4.3
5.4.4.4
5.4.4.4.1
5.4.4.4.2
5.4.4.4.3
5.4.4.4.4

Background................................................................................................................................................ 63
Organizations ............................................................................................................................................. 63
Survey Data: Automated Clearing House .................................................................................................. 64
ACH Transaction Classifications............................................................................................................... 64
Network vs. On-Us Transactions.......................................................................................................... 64
Originated vs. Received Transactions................................................................................................... 65
Inter- vs. Intra-Operator Transactions................................................................................................... 65
Debits vs. Credits.................................................................................................................................. 65

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

6
5.4.4.4.5
5.4.4.4.6

5.4.5

Electronic Benefit Transfer..................................................................................................................67

5.4.5.1
5.4.5.2
5.4.5.3

5.4.6

Emerging Payment Technology Companies ........................................................................................69

5.4.6.1

Returns.................................................................................................................................................. 65
SEC Codes............................................................................................................................................ 66

Background................................................................................................................................................ 69

APPENDICES ..................................................................................................................................................71
6.1
APPENDIX A – STATISTICAL METHODOLOGY FOR THE DFI CHECK STUDY ..............................................71
6.1.1
Sample Design .....................................................................................................................................71
6.1.1.1
6.1.1.2
6.1.1.3
6.1.1.4
6.1.1.5

6.1.2

Sample Weighting ................................................................................................................................74

6.1.2.1
6.1.2.2
6.1.2.3

6.1.3

Sampling Frame of the Financial Institutions ............................................................................................ 71
Stratification of Sampling Frame ............................................................................................................... 72
Sample Size and Sample Allocation .......................................................................................................... 72
Sample Selection........................................................................................................................................ 73
Selection Probabilities ............................................................................................................................... 74
Base Weights ............................................................................................................................................. 74
Nonresponse Adjustment ........................................................................................................................... 75
Ratio Estimation ........................................................................................................................................ 75

Estimation ............................................................................................................................................76

6.1.3.1
6.1.3.2
6.1.3.3
6.1.3.4
6.1.3.5
6.1.3.6
6.1.3.6.1
6.1.3.6.2

Estimates of Totals..................................................................................................................................... 76
Reliability of the Estimates ........................................................................................................................ 77
Variance Estimation of Estimates of Totals ............................................................................................... 78
Construction of Confidence Intervals ........................................................................................................ 79
Restratifying with the New PCD Data ....................................................................................................... 80
Imputation.................................................................................................................................................. 82
Imputation Methodology for Aggregate Data....................................................................................... 82
Imputation Methodology for Clearing Method Data ............................................................................ 83

6.2
APPENDIX B – STATISTICAL METHODOLOGY FOR THE CHECK SAMPLE STUDY ........................................83
6.2.1
CSS Sample Design..............................................................................................................................83
6.2.2
CSS Sample Weighting.........................................................................................................................85
6.2.3
CSS Estimates ......................................................................................................................................86
6.2.4
Random Sampling in the Field.............................................................................................................87
6.2.4.1
6.2.4.2
6.2.4.3
6.2.4.4
6.2.4.5
6.2.4.6
6.2.4.7

6.3
6.4
6.5

The Sampling Parameters Request Form ................................................................................................... 87
Master List of Random Checks.................................................................................................................. 88
Master List of Random Sequence Numbers............................................................................................... 88
Random Sampling Wizard ......................................................................................................................... 89
Customized Randomization Schemes ........................................................................................................ 89
Photo Retrieval Latitude ............................................................................................................................ 90
DFI-Selection of Sampled Checks ............................................................................................................. 90

APPENDIX C – THE DFI CHECK STUDY SURVEY INSTRUMENT .................................................................91
APPENDIX D – THE CHECK SAMPLE STUDY SURVEY INSTRUMENT (ANSWER SHEET) ..............................91
APPENDIX E – THE SAMPLING PARAMETERS REQUEST FORUM (THE SCREENER).....................................91

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

7
Table of Exhibits
Exhibit 1: Comparison of Check Volume Survey Results (1979 vs. 2001)................................................ 11
Exhibit 2: Distribution of Check Volume and Dollar Value by Clearing Method ..................................... 12
Exhibit 3: Estimated Distribution of Check Volume and Dollar Value by Payment Purpose.................... 15
Exhibit 4: Purpose of Check Payments by Consumers (Volume) .............................................................. 16
Exhibit 5: Purpose of Check Payments by Business/Government Entities (Volume) ................................ 17
Exhibit 6: Comparative Distribution: Electronic Payments Volume vs. Dollar Value............................... 19
Exhibit 7: Comparison of Non-Cash Payments Volumes (1979 vs. 2000)................................................. 21
Exhibit 8: Original Check Categorization Matrix....................................................................................... 41
Exhibit 9: Mix of Electronic Payment Instruments..................................................................................... 53
Exhibit 10: Private-Label Credit Card Volume Mix................................................................................... 59
Exhibit 11: Private-Label Credit Card: Average Transaction Size by Category ........................................ 60

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

8
Table of Tables
Table 1: Estimated Returned Check Volume & Value Compared to Paid Check Volume & Value.......... 12
Table 2: Estimated Distribution of Check Volume by Counterparty.......................................................... 13
Table 3: Estimated Distribution of Check Value by Counterparty ............................................................. 14
Table 4: Estimated Distribution of Check Volume by Counterparty and Purpose ..................................... 16
Table 5: Estimated Distribution of Check Value by Counterparty and Purpose ........................................ 17
Table 6: Total Estimated Volume and Dollar Value of Electronic Payments ............................................ 18
Table 7: Number of Institutions Sampled by Size Stratum (Original Design) ........................................... 27
Table 8: Survey Response by Type of Institution....................................................................................... 30
Table 9: Estimated Annual Check Payments Volume ................................................................................ 31
Table 10: Estimated Annual Check Payments Value ................................................................................. 31
Table 11: Estimated Annual Check Volume Distributed by Clearing Method........................................... 32
Table 12: Estimated Annual Check Value Distributed by Clearing Method.............................................. 32
Table 13: Estimated Annual Aggregate Outgoing Returns (Volume and Value)....................................... 32
Table 14: Stage One Sample Allocation – DFIs Sampled per Stratum ...................................................... 34
Table 15: Stage Two Sample Allocation – Checks Desired per Stratum (Design)..................................... 34
Table 16: Payments Classification Factors ................................................................................................. 38
Table 17: CSS Response Rate by Stratum .................................................................................................. 44
Table 18: Checks Surveyed for CSS by Stratum (Actual Response).......................................................... 45
Table 19: Response Rate (in Checks) as a Percentage of the Original Design ........................................... 45
Table 20: Estimated Distribution of Check Volume by Payer.................................................................... 46
Table 21: Estimated Distribution of Check Volume by Payee ................................................................... 46
Table 22: Estimated Distribution of Check Volume by Counterparty........................................................ 46
Table 23: Estimated Distribution of Check Volume by Counterparty and Purpose ................................... 47
Table 24: Estimated Distribution of Check Value by Payer ....................................................................... 47
Table 25: Estimated Distribution of Check Value by Payee....................................................................... 47
Table 26: Estimated Distribution of Check Value by Counterparty ........................................................... 48
Table 27: Estimated Distribution of Check Value by Counterparty and Purpose ...................................... 48
Table 28: Estimated Average Value per Check Category (Counterparty by Purpose) ............................... 48
Table 29: Distribution of Checks by Dollar Amount.................................................................................. 49
Table 30: Types of Transactions and Organizations Included in Study ..................................................... 51
Table 31: Study Participation Rate by Payment Instrument ....................................................................... 52
Table 32: Total Estimated Volume and Dollar Value of Electronic Payments .......................................... 53
Table 33: Types of Card Products............................................................................................................... 55
Table 34: Parties to Typical Credit-Card Transaction ................................................................................ 55
Table 35: General-Purpose Credit Card Volumes ...................................................................................... 57
Table 36: Private-Label Credit Card Volumes ........................................................................................... 59
Table 37: Online vs. Offline Debit Cards ................................................................................................... 60
Table 38: Online Debit Card Volumes ....................................................................................................... 62
Table 39: Offline Debit Card Volumes....................................................................................................... 62
Table 40: Automated Clearing House Volumes ......................................................................................... 64
Table 41: SEC Codes Included in ACH Aggregates .................................................................................. 66
Table 42: SEC Codes Excluded from ACH Aggregates............................................................................. 66
Table 43: ACH Transactions By SEC Code ............................................................................................... 66
Table 44: Electronic Benefit Transfer Volumes ......................................................................................... 69
Table 45: Number of Institutions Sampled by Size Stratum (Original Design) ......................................... 73
Table 46: Final Depository Financial Institution Check Study Sample Information.................................. 81
Table 47: Stage One Sample Allocation – DFIs Sampled per Stratum ...................................................... 84
Table 48: Stage Two Sample Allocation – Checks Desired per Stratum (Design)..................................... 85

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

Chapter 1

Executive Summary

1
1.1

Executive Summary
Introduction

In 2001, the Federal Reserve System made its first attempt to directly estimate the volume and
value of checks and other retail payments in more than 20 years. As the country's central bank,
the Federal Reserve undertook this research a) to document the size of the U.S. retail payments
system and b) to quantify the opportunity for future substitution of paper check payments by
electronic payments. Based on interactions with industry stakeholders before, during and since
the study, this effort has been welcomed by the industry.
Three studies – two examining the check payments market and one measuring electronic
payments – measure the size of the non-cash payments market and offer unique insights into the
dynamics of the U.S. retail payments system.
The results are used to provide a snapshot of the current U.S. retail payments market. The 2001
effort lays an essential foundation for continued research and trend analysis by the industry and
the Federal Reserve System.
The Federal Reserve's 2001 Retail Payments Research Project consisted of the following:
The Depository Financial Institution Check Study estimated the total annual volume
and value of check payments in the United States. Conducted with the help of
Global Concepts and Westat, The Depository Financial Institution (DFI) Check
Study collected data from a sample of depository financial institutions about paid
checks and returned checks during March-April, 2001. Using the two-month
sample, estimates for the entire industry were produced and are reported on an
annual basis.
The Check Sample Study complemented The DFI Check Study by characterizing check
payments according to type of payer, payee and purpose. Specifically, The Check
Sample Study (CSS) estimates who (consumer, business or government) writes
checks to whom (consumer, business or government) and why (remittance, pointof-sale, income or casual payments). This study, also conducted with Global
Concepts and Westat, surveyed depository financial institutions to estimate
national check usage.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

10
The Electronic Payment Instruments Study estimated the total volume and dollar value
of payments across each of six established electronic payment mechanisms.1 The
Electronic Payment Instruments Study (EPIS), conducted with the help of Dove
Consulting, estimates the transaction volume and value of credit cards (general
purpose and private label), debit cards (online and offline), ACH (automated
clearing house) payments and EBT (electronic benefits transfer) payments by
surveying clearing houses and payment processors through which these payment
volumes are cleared. The study estimates national transaction volumes for the
year 2000.
This report describes the results of each study, beginning with The DFI Check Study and ending
with The Electronic Payment Instruments Study.
The three studies paint a unique picture of the size and characteristics of the non-cash payments
market. They will serve as a benchmark against which the industry can measure its evolution and
its success in converting paper payments to electronic transactions.
1.2

Significant Results: The Depository Financial Institution Check Study

The DFI Check Study estimates that 42.5 billion checks are written annually in the U.S.
accounting for $39.3 trillion in payments.2
The DFI Check Study, which gathered paid check volume data for the months of March and
April 2001 from 1,256 U.S. financial institutions, is the first comprehensive check volume study
since the 1979 Federal Reserve Check Collection Study. That study estimated a national volume
of 32.8 billion checks. A comparison of the 1979 and 2001 studies shows that the number of
check payments has risen 30% from the 1979 estimate. This translates into an average compound
annual growth rate (CAGR) of 1.2% for check payments between 1979 and 2001.3

The Study also describes the usage levels of several emerging payment mechanisms (e.g., stored value cards,
Internet currencies, etc.).
2
Including 95% confidence intervals, the estimates are 42.5 billion checks +/- 1.59 billion checks and $39.3 trillion
+/- $2.46 trillion respectively. These estimates include 492 million Federal government checks and Postal Money
Orders for a total value of $313 billion that were not estimated by the survey but added to the final survey results.
3
This study provides no information about how growth rates may have varied from year to year.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

11
Exhibit 1: Comparison of Check Volume Survey Results (1979 vs. 2001)
45
40

Billions of Payments

35
30
25
20
15
10
5
1979

2001

The average value per check has increased 22%, from $757 in 1979 to $925 in 2001. This
increase in average value, however, can be attributed to inflation. The average value of an
inflation-adjusted check has actually dropped since 1979. In 2001 real dollars the 1979 estimate
would be approximately $1,584 compared to $925 today.
The drop in average dollar value per check payment likely reflects substitution by wire and ACH
payments for high dollar value checks.
1.2.1

Check Volume and Value by Clearing Methods

In addition to estimating national check payments volume and value, The DFI Check Study
gathered data on the methods used to clear and settle checks. Twenty-nine percent of all checks
are “on-us,” meaning the bank of first deposit for these items is also the institution on which the
checks are drawn. These 12.4 billion on-us checks ($14.3 trillion in value) are not cleared and
settled between financial institutions.
"On-us" checks represent a greater portion of check value than of check volume: 36% of total
U.S. check value, equivalent to $14.3 trillion, are “on-us” checks. In contrast, clearing houses,
same-day settlement and the Federal Reserve all represent smaller shares of check value than of
check volume.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

12
Exhibit 2: Distribution of Check Volume and Dollar Value by Clearing Method
Total Check Value: $39.3 trillion

Total Check Volume: 42.5 billion
Same Day
Settlement
6%

Other
5%

Clearing
Houses
18%

Treasury /Postal
Money Order
1%

Same Day
Settlement
4%

Other
6%

Clearing Houses
15%

Federal Reserve
Banks
38%

Federal Reserve
Banks
41%
On-Us
36%

On-Us
29%

1.2.2

Treasury /Postal
Money Order
1%

Returned Check Volume and Value

Another objective of The DFI Check Study was to collect volume and value estimates for
returned checks.4 The study found that 251 million of the 42.5 billion checks written annually are
returned. This equates to 0.6% of total check volume. The average value per returned check is
$701, which is $224 less than the average value across all check payments.5
Table 1: Estimated Returned Check Volume & Value Compared to Paid Check Volume & Value
Volume
Returned Checks
Total Checks

1.3

Value

Average Value

251 million

$176 billion

$701

42,508 million

$39,309 billion

$925

Significant Results: The Check Sample Study

Conducted as a complementary study to The DFI Check Study, The Check Sample Study
examined 28,877 randomly selected checks deposited at 149 depository financial institutions in
the U.S. The study categorized these checks by type of payer, payee and purpose. In terms of

The study did not attempt to quantify the volume or value of returns re-presented. Re-presentment of returns will
have a very slight inflationary effect on the aggregate volume/value of paid checks.
5
The $701 average value per return is also considerably lower than the $1,100 noted in a 1995 report to the
Congress.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

13
volume, results indicate that consumers are the predominant check writers, but businesses
receive the majority of checks.
The table that follows lists the estimated distribution of check payments across various
combinations of payer (the party making the payment) and payee (the party being paid). The
payer and payee together constitute the counterparty. The table illustrates that consumers write
over half of all checks, while businesses receive about half of all checks.
Table 2: Estimated Distribution of Check Volume by Counterparty
Check Volume by Counterparty with 95% confidence interval (for totals only)
Payee
Business or
Consumer Business Government Government
Payer
Consumer
11.2%
33.8%
1.7%
2.7%
Business
14.6%
15.0%
0.8%
1.0%
Government
2.5%
0.7%
0.2%
0.0%
Business or Govt6
0.6%
0.4%
0.0%
0.1%
Unknown7
0.1%
0.2%
0.0%
0.0%
29.1%
50.1%
2.7%
3.8%
Total
+/- 4.4% +/- 3.2%
+/- 0.5%
+/- 0.6%

Unknown
1.6%
0.9%
0.1%
0.1%
11.6%
14.3%
+/- 2.5%

Total
50.9% +/- 2.2%
32.3% +/- 2.1%
3.5% +/- 0.7%
1.3% +/- 0.1%
12.0% +/- 2.6%
100.0%

In contrast to the check volume analysis, consumer-written checks account for only 19% of the
total value of check payments, while businesses write checks for 62% of total check value. In
terms of value, businesses are both the heaviest writers and receivers of check payments. Taken
alone, business-to-business checks account for over 40% of the total value of check payments.

The Business or Government (BG) category includes checks for which the payer or payee is clearly not a consumer
but cannot be distinctly categorized as a business or government. Business and governments are often grouped
together in the study because their payment behavior is very similar and because governments represent a small
portion of check volume compared to businesses.
7
The “unknown” category in Table 2 and Table 3 includes both those items for which the survey was filled out with
inconsistent data (approximately 2%) and those items for which no payer, payee and/or purpose (approximately
10%) was identified. The “unknown” category may include specific segments of check payments that are difficult to
categorize; for example, a number of checks written by sole proprietors or very small businesses (i.e., checks that
have many characteristics of consumer payers and business payers) may have been categorized as “unknown” for
payer. Checks for which it is unclear whether the payee is a consumer or small business are likely to be categorized
as “unknown” for payee and for purpose.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

14
Table 3: Estimated Distribution of Check Value by Counterparty
Check Value by Counterparty with 95% confidence interval (for totals only)
Payee
Business or
Consumer Business Government Government Unknown
Payer
Consumer
6.7%
10.2%
0.6%
0.7%
0.9%
Business
14.5%
42.7%
0.5%
2.9%
0.9%
Government
1.5%
1.9%
0.3%
0.0%
0.0%
Business or Govt
0.4%
0.5%
0.0%
0.1%
0.5%
Unknown
0.1%
0.2%
0.0%
0.0%
13.6%
23.3%
55.5%
1.5%
3.8%
15.9%
Total
+/- 4.0%
+/- 4.3%
+/- 0.6%
+/- 1.9%
+/- 2.9%

1.3.1

Total
19.2% +/- 2.1%
61.6% +/- 3.5%
+/- 1.1%
3.8%
+/- 0.5%
1.5%
+/- 2.9%
13.9%
100.0%

Categorization of Check Payments by Purpose

The study categorized check payments into six purpose categories: Casual, Income, Remittance,
Point of Sale (POS), Remit/POS, and Unknown. These categories are defined as follows:

Casual: Payment from one individual to another. By definition, all consumer-toconsumer payments, therefore, are categorized as Casual. This category is likely to
include some payments to and from individuals acting as small businesses, such as rent
payments written to individuals.
Income – Payment to an individual from either a business or government entity. By
definition all business-to-consumer or government-to-consumer payments, therefore, are
categorized as Income. This category includes – for example – payroll, pension, rebates,
expense reimbursement and investment disbursement checks.
Remittance –Payments from any type of payer to either a business or government payee
that does not occur at the point of sale. The types of remittance payments include: regular
recurring remittance payments such as bill payments, non-recurring bill payments such as
payments to doctors, plumbers, etc., and commercial remittance payments between
businesses and/or government organizations.
Point of Sale (POS) – Payments from any type of payer to either a business or
government payee that occurs in any of the following environments: storefront, over-thecounter retail remittance, such as a telecom bill paid at the local office, MOTO (mail
order/telephone order, e.g., catalog retailers), Internet, mobile POS (such as check
payments occurring at home to repairmen), C.O.D., and vending.8
Remit/POS – Payments made to business or government payees that The Check Sample
Study was unable to categorize as either distinctly remittance or POS.
Unknown – Payments for which The Check Sample Study could not determine a purpose
or for which the survey respondent may have completed the questionnaire incorrectly.

As shown in the exhibit below, remittance and POS check payments combined represent over
half of all check payments volume (57%). Income and casual check payments (those payments
8

Internet, MOTO (mail/telephone order), and vending transactions fall into the POS category, but they do not apply
to check payments. The categorization was designed to describe all potential payment mechanisms – not just check
– for the POS purpose category.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

15
received by consumers) represent about 30% of all check payments by volume, while Unknown
checks account for the remainder.
The distribution of check payments by dollar value demonstrates that POS checks, while a
significant portion of total volume (19%) make up only 9% in terms of the total value of checks.
Conversely, the Remittance/POS category accounts for a significantly greater portion (25%) of
check payments value than of volume (12%). This is due to a relatively small number of highvalue checks between business or government payers and payees for which the purpose (either
Remittance or POS) could not be clearly determined.
Exhibit 3: Estimated Distribution of Check Volume and Dollar Value by Payment Purpose
Check Payments Volume by Purpose
Unknown
14%

Income
18%

Remit/POS
12%

POS
19%

Casual
11%

Rem
26%

Check Payments Value by Purpose
Unknown
16%

Income
16%
Casual
7%

Remit/POS
25%

POS
9%

Rem
27%

Check payments can be categorized into eight Payer-Payee-Purpose subcategories (excluding
Unknown) that describe the distribution of check payments.9 As shown in the chart below, the
largest segments of check payments are business and/or government income payments to
consumers (17.8%) and consumer remittance payments to business and/or governments (17.7%).
Table 4 illustrates that payments categorized as Remit/POS are roughly split between those
written by consumers (6.4%) and those written by businesses or government organizations
(5.5%). Remit/POS payments written by businesses or government organizations are more likely
to be remittances. Those written by consumers are more difficult to surmise. This segment likely
includes POS payments from the home to contractors and remittance payments to small
businesses that endorse checks in a similar way to small retail merchants or consumers.

The intersection of Payer, Payee and Purpose actually allows for 20 cells (excluding "unknown"), but only 8 cells
are by definition appropriate for analysis. For example, consumers do not receive POS payments.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

16
Table 4: Estimated Distribution of Check Volume by Counterparty and Purpose
Distribution Includes 95% confidence interval (% All Check Payments)

Purpose
Income
Casual
Remittance
POS
Remit/POS
Unknown
Total

C2C

(+/-)

C2BG

(+/-)

Counterparty
BG2C (+/-) BG2BG
17.8% 2.5%

(+/-)

Unknown10

(+/-)

0.7%
0.8%
0.9%

0.1%
0.1%
0.1%
14.5%
14.7%

0.1%
0.0%
0.0%
2.5%

11.2% 1.9%
17.7% 1.9%
14.1% 2.0%
6.4% 0.9%
11.2%

38.1%

7.9%
4.9%
5.5%
17.8%

18.3%

Total
17.8%
11.2%
25.7%
19.0%
11.9%
14.5%
100.0%

Of the estimated 21.6 billion checks written by consumers, 22% are for casual payments.
Exhibit 4: Purpose of Check Payments by Consumers (Volume)
21.6 Billion Consumer Checks
Casual
22%

Remit/POS
13%

Unknown
3%

Remittance
34%

POS
28%

About half (48%) of the estimated 15.8 billion checks written by businesses or government
organizations are income payments to consumers. Remittance and POS payments make up the
other half.

Checks for which the counterparty relationship is unknown are any checks for which either payer or payee cannot
be determined. Therefore, some checks for which the payer is known (e.g., consumer) have been categorized as
"unknown" for the purpose of this table, because the payee type could not be determined. The sum of all C2C and
C2BG checks, therefore, does not equal the sum of all consumer checks. Nor do BG2C and BG2BG account for all
checks written by business or government payers.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

17
Exhibit 5: Purpose of Check Payments by Business/Government Entities (Volume)

15.8 Billion Bus/Gov Checks

Remit/POS
15%

Unknown
3%

POS
13%

Income
48%
Remittance
21%

Nearly half (49%) of the value of check payments is concentrated into a single counterparty
relationship – BG2BG payments (payments from business or government payers to business or
government payees), with 21% going toward remittance payments, 21% to remittance or POS
payments and 7% for POS payments. The only other segment with more than 7% of the total
value of check payments is business/government income payments to consumers.
Table 5: Estimated Distribution of Check Value by Counterparty and Purpose
Distribution Includes 95% confidence interval (% All Check Payments Value)

Purpose
Income
Casual
Remittance
POS
Remit/POS
Unknown
Total

1.4

C2C

(+/-)

6.7%

1.2%

6.7%

C2BG

(+/-)

Counterparty
BG2C
(+/-) BG2BG
16.5% 3.6%

6.2%
1.7%
3.7%

1.5%
0.4%
0.9%

21.1%
7.1%
20.8%

11.5%

16.5%

(+/-) Unknown

4.9%
2.2%
4.6%

49.0%

0.1%
0.0%
0.1%
16.1%
16.2%

(+/-)

0.1%
0.0%
0.1%
2.9%

Total
16.5%
6.7%
27.4%
8.8%
24.6%
16.1%
100.0%

Significant Findings: Electronic Payment Instruments Study

Conducted independently of the two check studies above, The Electronic Payment Instruments
Study (EPIS) estimated the volume and dollar value of electronic payments. The study estimates
that during calendar year 2000, 29.5 billion electronic payments were originated in the United
States with a value of $7.3 trillion.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

18
The study counted electronic payment transactions from consumers, businesses and government
entities in three primary areas: those used by buyers of goods or services; those used on the
‘back-end’ to effect final settlement for purchase transactions; and those used by employers, state
agencies and others for disbursements of income payments. The Study excluded non-purchase
transactions such as ATM and settlement transactions, which are not considered to be direct
substitutes for paper checks. Fedwire, CHIPS, and automated clearing house (ACH) transactions
for Cash Concentration or Disbursements (i.e. CCD) were all excluded from the study.
The Study was conducted by surveying the leading electronic payment processors. The research
team identified and surveyed 118 organizations that could collectively provide data on all
electronic payment transactions in the United States. Organizations representing 94% of all
estimated payment value participated in the survey. Data for the remaining organizations were
estimated.
Table 6: Total Estimated Volume and Dollar Value of Electronic Payments

Electronic Payment Instrument
General purpose credit cards11
12

Private label credit cards

Offline debit (signature-based)
14

Online debit (PIN-based)

Transaction
Volume
(Millions)

Dollar
Volume
($Millions)

Average
Payment
Value

12,300.2

$1,072,555

$87.20

2,748.6

$162,819

$59.24

5,268.6

$209,980

$39.85

3,010.4

$138,151

$45.89

5,622.0

$5,674,851

$1,009.40

537.7

$13,744

$25.56

29,487.5

$7,272,100

$246.62

Automated Clearing House (ACH)

Electronic Benefits Transfer (EBT)
Total

General-purpose credit cards include co-branded credit cards, charge cards, co-branded charge cards
secured credit cards, T&E cards, commercial cards (including business, corporate, and purchasing), and new
payment technologies that route transactions through the card associations' networks.
12
Private-label credit card programs include those run by individual retailers or gas companies, third-party fleet-card
issuers and third-party receivable owners.
13
Offline debit refers to debit card transactions that require the customer's signature (rather PIN) as a means of
authentication. Visa and MasterCard have the only two networks for offline debit transactions. Visa’s offline debit
statistics also include the new hybrid online/offline debit card offered by Visa, the Visa Check Card II.
14
Online debit card transactions require the customer to enter a four-digit PIN (personal identification number) as a
means of authentication (as opposed to signing a sales receipt). Online debit card transactions are originated, cleared
and settled over EFT networks (the same networks that process ATM transactions) as opposed to Visa or
MasterCard's networks.
15
The ACH is an electronic payments network that allows credits and debits to be processed between financial
institutions. ACH transactions between financial institutions are processed by one of four ACH operators, who were
surveyed. ACH transactions that are On-Us (a DFI is both the originator and receiver of a transaction) have been
estimated using data from NACHA (National Automated Clearing House Association).
16
EBT is an electronic system that allows a recipient to authorize transfer of his/her government benefits from a
Federal account to a retailer account to pay for products received. EBT is currently being used in many states to
issue food stamps and other benefits.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

19
As shown in the exhibit below, the majority (51%) of electronic payment transactions were made
using credit cards, but 78% of payment dollars were handled through the ACH.
Exhibit 6: Comparative Distribution: Electronic Payments Volume vs. Dollar Value
Electronic Payments Volume by Payment Instrument

ACH
19%

EBT
2%

1.4.1

ACH
78%

General Purpose
Credit Cards
42%

Online Debit Cards
10%

Offline Debit
Cards
18%

Electronic Payments Value by Payment Instrument

Private Label
Credit Cards
9%

EBT
0%
General Purpose
Credit Cards
15%

Online Debit Cards
2%

Offline Debit
Cards
3%

Private Label
Credit Cards
2%

General Purpose and Private Label Credit Card

On a transaction volume basis, general purpose and private label credit cards were the most
common electronic payment instrument used in the U.S. during the Year 2000: 15.0 billion
transactions were originated with a value of $1,235 billion. The average transaction size for
general purpose credit cards was much larger than that of private label cards: $87.20 vs. $59.24.
Credit cards accounted for 51% of all electronic payment transactions and 17% of the dollar
value. Eighty-two percent of credit card transactions and 87% of transaction value came from
general purpose credit cards.
1.4.2

Online and Offline Debit Cards

Following credit cards, debit cards represented the second most common form of electronic
payment, accounting for 8.3 billion transactions and a dollar value of $348 billion in 2000. On
average, each debit transaction was $42, compared with $87 for the average general purpose
credit card transaction. In 2000, 64% of transactions and 60% of the value was contributed by
offline debit (i.e., signature-based); 36% of transactions and 40% of value was from online debit
(i.e., PIN-based).
1.4.3

ACH

Although ACH was the third most commonly used electronic payment instrument with 5.6
billion transactions, it dominates on a dollar value basis accounting for 78% of the monetary
value. The average transaction volume was more than 11 times larger than that of general
purpose credit card transactions ($1,009 vs. $87).

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

20
1.4.4

EBT

EBT volume has increased dramatically due to initiatives at the federal level and significant
efforts by state governments to electronify both food stamps and cash assistance payments
during the 1990s. Nevertheless, EBT was the smallest volume payment instrument with 500
million transactions and $13.7 billion in value. Note that the “EBT” category in this study refers
to consumer payments using EBT. Government disbursements to financial institutions that hold
EBT funds and those institutions’ reimbursements to merchants for EBT sales are included in the
ACH category.
1.4.5

Emerging Payments

The survey of emerging payments involved companies that provide services in such markets as
electronic bill payment, person-to-person payments, stored value, Internet currencies and other
emerging technologies. In general, emerging payment volumes for the payment instruments
studied were quite small in 2000. Organizations participating in the survey reported 76.2 million
transactions involving $12.7 billion. However, these numbers represent only a small portion of
the total emerging payments. Many organizations did not respond to the survey because they
were very new or they were in a very competitive market and did not want to reveal their data.
The Study did not attempt to estimate the volumes for non-respondents. Several categories
within the emerging payments group will be important to watch in the coming years, especially
person-to-person payments.17
1.5

Non-Cash Payments Market Then vs. Now

The Federal Reserve's 1979 check collections study was the last authoritative study of the U.S.
check payments market. The study did not estimate electronic payments volumes comparable to
the 2001 effort, but review of available data suggests a 1979 market of approximately 5-6 billion
electronic payments.18
Since 1979, the total number of non-cash retail payments has nearly doubled from approximately
38 billion to 72 billion. This translates into an average compound annual growth rate (CAGR) of
2.9% for non-cash retail payments from 1979 to 2001.19 ACH, credit card, debit card, and EBT
payments have led this growth: taken together these four electronic retail payment types have
grown at an average CAGR of approximately 8%, increasing from an estimated 5-6 billion in
1979 to 29.5 billion, according to the results of the Electronic Payment Instruments Study.
Check volume on the whole has increased by 30% since the 1979 estimate of 32.8 billion checks.
Although checks remain the dominant form of non-cash payment, over the last 20 years, their
proportion of the total payments market has declined considerably. Despite overall growth in
17

Emerging payments settle via existing payment systems for which this study estimates annual volumes.
The retail electronic payments volume in 1979 was attributable entirely to credit card and ACH transactions. The
estimated volume of 5-6 billion payments is based on extrapolation from a study by the Federal Reserve Bank of
Cleveland for its 1977 annual report. That report estimated 5 billion credit card transactions and 12 million ACH
transactions in 1976-77. Available data from general purpose credit/charge card companies and secondary sources
support a conservative estimate of 5-6 billion electronic payments in 1979.
19
The 2001 research effort marks the first time since 1979 that the Federal Reserve System conducted research to
estimate national check volume. The characteristics of the growth curve between 1979 and 2001 for paper or
electronic payments are unknown; the CAGR represents the average growth between the 1979 and 2001 data.
18

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

21
volume, checks have declined from approximately 85% of non-cash payments in 1979 to 59%
today.
Exhibit 7: Comparison of Non-Cash Payments Volumes (1979 vs. 2000)
Growth of Non-Cash Retail Payments: Check Vs. Other
80
70

Billions of Payments

ACH, CC, DC, EBT
Checks

50
40
30
20
10
1979

1.6

2000

Implications and Observations

As the Federal Reserve, like many other payment processors, looks for ways to make the
payments system more efficient, it is vital to understand where opportunities exist for migrating
check payments to electronics.
The data from this research show that remittance and point-of-sale payments written by
consumers offer the most significant opportunities for substitution, as these are the largest
categories of checks written today. This implies that the ACH and credit and debit cards are
poised for meaningful growth in the near term.
Given the large number of checks still being written in the United States and the increased usage
of electronic forms of payment, businesses and financial institutions are going to have to
maintain multiple channels for the foreseeable future. While checks, we believe, will account for
a decreasing portion of total payments, they will continue to be around for some time to come.
Despite an annual volume of 42.5 billion checks, it appears that Americans are changing their
historical, conservative use of payments. The fact that 30 billion electronic payments were
initiated in 2000 indicates clear acceptance by consumers and businesses. And given that debit
and credit card payments and ACH transactions collectively have grown exponentially in the last
20 years – some 500 percent since 1979 – these electronic forms of payment will become more
prevalent and increasingly a requirement of doing business for U.S. companies and financial
institutions.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

22
The data from the study are important because they offer factual evidence to financial
institutions, the financial services industry and the Federal Reserve System on the volume and
value of payments. From this, industry stakeholders may make inferences about the migration of
the payments system and where prudent opportunities for investments in payments system
technology exist. Going forward, the Federal Reserve plans to repeat this research and establish a
trend line that will enable both industry stakeholders and the Fed to measure the progression of
the payments system and the migration of paper payments to electronics. This study is just a
picture – a snapshot – of the continuing evolution of the payments system.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

Chapter 2
Introduction

Introduction

This report details the methodologies and findings of each of three research efforts performed as
part of The Retail Payments Research Project:

2.1

The Depository Financial Institution Check Study

Chapter 3

The Check Sample Study

Chapter 4

The Electronic Payment Instruments Study

Chapter 5

Survey Methodologies

Each of the three studies used a discrete methodology, but a distinction should also be made
between the two studies of check payments and the study of electronic payments:
•

Both The Depository Financial Institution (DFI) Check Study and The Check Sample
Study were sample surveys. They relied on standard statistical techniques in order to
estimate the size and characteristics of the check payments market by surveying a
representative sample of the whole. For The DFI Check Study – which estimated the
national annual volume and dollar value of check payments – this meant surveying a
representative random sample of insured depository financial institutions with regard to
their paid check volume. The Check Sample Study – which characterizes the types of
check payments being made – required that a representative random sample of checks be
surveyed.
As sample surveys The DFI Check Study and The Check Sample Study are subject to
sampling error. For this reason tables in this report include error estimates alongside
many of the point estimates. The error estimate, or margin of error, reflects an interval
within which there is a 95% level of confidence about the estimate. The 95% confidence
intervals are based on sampling error alone.20

•

The Electronic Payment Instruments Study did not rely on a statistical sampling
methodology. Rather, the study took a census of clearinghouses and payment processors
who clear and settle the vast majority of electronic payments. While not all processors

There are non-sampling error factors not accounted for in these error estimates. Non-sampling errors include
biases due to inaccurate reporting, processing and measurement, as well as error due to nonresponse and incomplete
reporting. These types of errors cannot be measured readily. However, to the extent possible each error has been
minimized through the procedures used for data collection, editing, quality control and nonresponse adjustment.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

24
chose to respond to the study, estimation of uncounted volume was minimized by strong
participation and thorough validation. We believe the results to be highly accurate
estimates of the total annual volume and dollar value of electronic payments in the United
States.
Combined the three studies allow the first authoritative account of both the size and
characteristics of non-cash retail payments in the United States in more than 20 years.
The Federal Reserve wishes to recognize the efforts of its contractors – Global Concepts, Dove
Consulting and Westat. We appreciate the quality of their execution of these studies.
We also extend our deepest gratitude to the many financial institutions and payment processors
that contributed data for the studies. Despite the considerable effort required of some institutions,
participation in the three studies was outstanding. The industry's commitment signifies clearly
the importance of understanding the dynamics of the market in which we operate. The Federal
Reserve System is truly grateful for the time and effort committed by all participants in The
Retail Payments Research Project.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

Chapter 3
DFI Check Study

The Depository Financial Institution Check Study

The survey research team of the Federal Reserve, Global Concepts and Westat conducted a
stratified random sample survey of depository financial institutions (DFIs) to estimate the total
volume of check payments in the U.S. over a 12-month period.21 Sample data were collected for
the months of March and April 2001. Results were extrapolated to the sample universe and
annualized. Participant DFIs were asked to provide the total volume and dollar value of paid
checks (i.e., On-Us checks) and (outgoing) returned check processed during each of the two
months.
The main objective of the study was to estimate the annual volume and value of check payments
in the U.S. The estimate includes data about the volume of checks paid by commercial banks,
thrifts and credit unions as well as the annual aggregate volume and dollar value for all
institutions.
3.1

Methodology

The DFI Check Study received check volume and value data from 1,256 financial institutions
during March and April 2001. The methodology discussion below reviews the study's
methodology at a relatively high level. For a more detailed description, see section 6.1 Statistical
Methodology for The DFI Check Study.
3.1.1

Sample Design

The study of paid check volume was conducted as a sample survey of DFIs. DFIs were stratified
before sampling, first by type of institutions and then by size. The three primary strata (by type
of institution) were commercial banks (CMB), credit unions (CUS) and thrifts (THR).22 The next
level of stratification was carried out on the basis of size where the measure of size was public
checkable deposits (PCD). The size stratification was based on the PCD value at the highest
institutional level (i.e., holding company if applicable). The sampling unit, therefore, was the
DFI at its highest institutional level (e.g., holding company) and the data were collected for all
the institutions owned by the sampled DFI.

A design variable called Public Checkable Deposits (PCD) was used to stratify the sample. These are all
checkable deposits held by a DFI that are not the deposits of other DFIs or the federal government. Checkable
deposits were believed to be a better indicator of check volume than total assets or total deposits. The sample was
also stratified according to type of institution – Commercial Bank, Thrift or Credit Union. See section 3.1.1 or
section 6.1.1 for more detail.
22
Thrifts include savings banks and savings and loan associations.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

26
3.1.1.1 Sampling Frame of the Financial Institutions

The sampling frame was constructed from files supplied by the Federal Reserve Board of
Governors. The frame represented the population of insured depository financial institutions in
the United States, which includes U.S. branches of foreign owned institutions. Only institutions
with checkable deposits above $100,000 were included in the frame. It is possible that a bank
holding company could have no checkable deposits, in which case it was eliminated from the
frame.
More specifically, the frame consisted of:
•
•
•

6,846 commercial banks and bank holding companies, plus 6 "anomalous banks"
6,551 credit unions,
1,293 thrifts.

The six anomalous banks were identified and surveyed as a certainty stratum, because their paid
check volume was known to be poorly correlated to PCD. Relatively speaking, these were small
banks (low PCD value) that process a high volume of low dollar value rebate checks. These
institutions were surveyed as a certainty stratum to avoid the risk of selecting them as part of a
random sample. They are not representative of most other institutions their size. Their data
would skew the results of a national estimate if it were extrapolated to estimate volumes
processed by non-sampled banks.
3.1.1.2 Certainty Strata

The study was designed to account for as much volume/value as possible and to estimate the rest
by surveying a representative random sample of DFIs. Because the largest institutions were
assumed to account for the majority of total paid check volume and dollar value, they were
sampled with certainty – i.e., all of them were included in the sample. Random sampling was
conducted to select institutions from the other strata.
3.1.1.3 Sample Size and Sample Allocation

The sample size of 2,339 DFIs was based on the following assumptions for each of the primary
strata (CMB, CUS and THR), and the aggregates of these strata:
•
•

Expected response rate of 65 percent.
Target of +/- 5% margin of error with a 95% level of confidence.

The total sample of 2,339 institutions was allocated across 14 design strata defined by type of
institution and size (plus one stratum of anomalous banks).
The table below gives the number of institutions in the sample frame and the number sampled
within each sampling stratum.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

27
Table 7: Number of Institutions Sampled by Size Stratum (Original Design)23
Primary
Stratum
Commercial
Banks

Credit
Unions

Thrifts

Size
Stratum

Number of
Institutions

1
2
3
4
5
6
All
1
2
3
4
5
All
1
2
3
All

TOTAL

3.1.2

204
329
845
1,408
2,036
2,008
6,830
104
344
723
1,742
3,199
6,112
40
347
850
1,237
14,179

Number
Sampled
204
329
336
300
242
156
1,567
104
181
134
112
69
600
40
114
44
198
2,365

Sample Weighting

Survey responses from the respondents were inflated to obtain estimates for the entire population
using a ratio estimation technique. In short, the ratio (i.e., sample weight) for each institution was
computed as the total PCD for its stratum divided by the PCD for the respondent institution. The
estimates for each stratum were then summed to produce a national estimate.
3.1.3

Estimation

The sampling and weighting methods above, combined with common imputation techniques,
allow the study's results to estimate the total annual volume and dollar value of the following:
•
•
•
•
3.1.4

Paid checks in the U.S.
Paid checks in the U.S. by type of institution.
Paid checks in the U.S. according to their clearing method.
Returned checks (i.e., outgoing returns).
Designing the Survey Instrument

Instrument development was an iterative process between Global Concepts and the Federal
Reserve. The final survey requested data from three major categories: paid checks, outgoing
returns, and routing transit numbers.
23

This table represents the original design of the sample. During the survey period a number of DFIs merged with
other institutions or were re-stratified for various reasons. The final sample size was 2,339 DFIs allocated as
follows: 1,547 commercial banks, 599 credit unions, 187 thrifts and 6 anomalous banks. See table 8 for a summary
or table 46 for details of the final stratification.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

28
3.1.4.1 Paid Checks

In designing the survey methodology, Global Concepts and the Federal Reserve decided to
survey paid checks. Check volume is prone to double counting, because multiple parties may
handle a single item in the check clearing process (e.g., bank of first deposit, payor bank,
correspondents). Because a check ultimately can be paid by only one DFI, however – the payor
bank – a survey of paid checks was deemed most likely to produce an accurate estimate of
national check volume.
Some DFIs do not process checks in-house. For example, many DFIs outsource check processing
either to other DFIs or to third-party check processors. For this reason, the questionnaire stressed
that the surveyed DFI was to report all checks paid by their institution, even if the checks were
processed by a third party.24 Conversely, DFIs that perform correspondent processing were asked
not to count checks they process on behalf of other DFIs.
3.1.4.2 Outgoing Returns

The survey instrument requested the volume and dollar value of all outgoing returns during the
survey period. These are checks presented to but returned unpaid by the respondent institution.
By surveying outgoing, as opposed to incoming, returns, the study avoided double-counting
returns handled by multiple institutions in much the same way as the survey of paid checks
avoids double counting presented check volume/value.
The survey did not attempt to distinguish the volume of re-presented returns. These are checks
returned once by the payor bank and presented a second time by the bank of first deposit for
payment. This volume has a small inflationary impact on the overall estimate of total paid check
volume.
3.1.4.3 Routing Transit Numbers (RTs)

In addition to surveying check volume and dollar value, the survey instrument requested that
each respondent report the RTs that correspond to all volume/value reported. The RT survey
provided a means to validate the identities of respondent DFIs and to gauge the reasonableness
of DFI-reported data. The RTs were used to compare respondent data to Federal Reserve data
about presentment volumes to those RTs.
3.1.4.4 Pretesting the Survey Instrument

The survey instrument was pretested with the help of both large and small financial institutions.
These institutions provided high quality feedback on the clarity of the survey instrument and the
relative difficulty in providing the survey data.
3.1.5

The Survey Instrument

Financial institution respondents reported value and volume data for the reference period (March
and April 2001) either via the paper survey that was mailed to them or on a secure Web site
using a user I.D. and a password provided to each institution. In addition to aggregate paid check
volume and value data, the survey also requested the responding companies to report

When necessary, the outsourcer was also included in the survey.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

29
components of the aggregate volume and value data for each month, breaking their total volume
and value down into five possible categories:
•
•
•
•
•

Inclearings from the Federal Reserve,
Inclearings from all Clearing Houses or local exchanges,
Inclearings via Same-day Settlement,
Inclearings from Other channels, and
On-Us deposits (i.e., those deposits for which the paying institution was the bank of first
deposit).

Responding institutions were also asked to report the volume and value of outgoing returns for
each month. Finally, each responding institution was requested to report active Routing Transit
numbers. A copy of The Depository Financial Institution Check Study survey instrument can be
found in Appendix C.
3.1.6

Reporting Period: March 1, 2001 - April 30, 2001

The survey research team chose a two-month survey period to mitigate any effect of an
aberration in check volume or value for any given month. March and April, 2001 were chosen,
because they were sufficiently representative without an unusual number of processing days. The
research plan called for adjusting the two-month data to an annual estimate of check volume and
value. The research team decided on a multiplication factor of 6 to annualize the two-month
data.25
While April is the end of the annual filing period for most personal income tax returns, this does
not have a significant effect on the overall estimates. Even without a seasonal adjustment, the
research team does not believe April's tax payment and refund volume would have a significant
impact on the overall estimate. Federal refund checks are paid by the U.S. Treasury and were,
therefore, not counted by the survey of depository financial institutions. Estimated annual federal
check volume and value (based on recent historical data) have been added to the national
estimate after survey results were extrapolated to the industry and annualized (see Table 9 and
Table 10). State refund checks, however, were counted through the survey of DFIs.
3.1.7

Data Collection

The data collection strategy was based on the Dillman Total Design Method (1978), which uses a
combination of respondent notifications, survey distribution and reminders to achieve the desired
response rate. This study used a modified form of the methodology that included multimode
contacts, that is, letter and phone calls, a different strategy for the top commercial banks certainty
stratum, and a choice of response modes, that is, letter, fax or website.
Extensive data scrubbing and follow-up were performed to verify data and to correct errors.

Historical processing volume data from the Federal Reserve implies that an annualization factor of 6 for the
combined volume of March and April is reasonable. (This reflects the conclusion that the reference period does not
exhibit either unusually high or low volume.) This factor assumes that there are no seasonal fluctuations in Federal
Reserve share of total industry volume.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

30
3.2
3.2.1

Results and Analysis
Survey Response

Fifty-four percent of sampled institutions (including 6 of 6 "anomalous banks") responded to the
survey. Credit Unions responded at the highest rate (57%).
Table 8: Survey Response by Type of Institution
SAMPLE

Sample
Frame Size

Institutions
Responding

Response
Rate

Commercial Banks

6,846

1,547

810

52%

Credit Unions

6,551

599

343

57%

Thrifts

1,293

187

52%

100%

14,696

2,339

1,256

54%

Anomalous Banks
All Institutions

3.2.2

Sample
Size

Aggregate Paid Check Volume and Value

The table that follows documents the survey's primary output – a 12-month estimate of the total
volume and value of check payments in the U.S. The 95% confidence interval is included for the
total and the breakdown of the total by the primary stratification (type of institution). The
aggregate total includes two additional data points that were not collected through the survey:
U.S. treasury checks and postal money orders (PMO). This volume (492 million items) and value
($313 billion) have been added to the estimated volume of checks paid by DFIs. Because
treasury checks and PMO volume are processed by the Federal Reserve, the volume and value
figures are known and include no error estimate.
Treasury checks and PMO volume and dollar value from the year 2000 were assumed to provide
the most reliable recent estimate of checks paid by the Federal Reserve, because Treasury check
volume in 2001 (or March-April 2001) was expected to produce an extraordinarily high estimate
– millions of tax refund checks were mailed to U.S. households in 2001.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

31
Table 9: Estimated Annual Check Payments Volume
Total Check Payments

95% Confidence Interval

U.S. Market

42.508

billion

(+/-) 1.593

billion

Commercial Banks
Credit Unions
Thrifts
Anomalous Banks

33.006
4.745
3.991
0.275

billion
billion
billion
billion

(+/-) 1.495
(+/-) 0.269
(+/-) 0.480
N/A

billion
billion
billion

0.262
0.230

billion
billion

U.S. Treasury Checks
Postal Money Orders

N/A
N/A

The estimated annual dollar value of check payments is summarized in the table below. The
aggregate estimate and the estimated breakdown by type of institution include a corresponding
95% confidence interval for each estimate.
Table 10: Estimated Annual Check Payments Value
Value of Check Payments

95% Confidence Interval

U.S. Market

$ 39.309 trillion

(+/-) $2.456 trillion

Commercial Banks
Credit Unions
Thrifts
Anomalous Banks

$ 36.549
$ 0.883
$ 1.552
$ 0.011

(+/-) $2.449 trillion
(+/-) $0.041 trillion
(+/-) $0.178 trillion
N/A

trillion
trillion
trillion
trillion

U.S. Treasury Checks $ 0.283 trillion
Postal Money Orders $ 0.030 trillion

3.2.3

N/A
N/A

Check Volume and Value by Clearing Method

Financial institutions were asked to report not only their total paid check volume and dollar
value, but also the distribution of that volume/value by presentment method. Specifically, DFIs
were asked to report volume/value received from the Federal Reserve, clearing houses (CH),
same-day settlement (SDS) and other clearing methods; as well as any "on-us" volume.26 The
results of the survey's clearing methods data are provided in the tables below. Each estimate
includes its corresponding 95% confidence interval. The figures exclude 492 million Treasury
checks and postal money orders for a value of $313 billion.

The category Other was used to allow DFIs to allocate volume they could not accurately label as either presented
by the Federal Reserve, a clearinghouse, via same-day settlement or deposited on-us.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

32
Table 11: Estimated Annual Check Volume Distributed by Clearing Method
Cleared via

Total Check Payments

95% Confidence Interval

FRB

17.223billion

(+/-)

N/A

ON-US

12.393billion

(+/-)

1.415billion

7.618billion

(+/-)

1.174billion

SDS

2.657billion

(+/-)

0.308billion

OTHER

2.125billion

(+/-)

0.522billion

(+ 0.492 billion Treasury checks and PMO)

Table 12: Estimated Annual Check Value Distributed by Clearing Method
Cleared via

Value of Check Payments

95% Confidence Interval

FRB

$ 14.639 trillion

(+/-)

N/A

ON-US

$ 14.286 trillion

(+/-) $ 1.611 trillion

$ 6.056 trillion

(+/-) $ 1.203 trillion

SDS

$ 1.637 trillion

(+/-) $ 0.277 trillion

OTHER

$ 2.377 trillion

(+/-) $ 0.827 trillion

(+ $ 0.313 trillion in Treasury checks and PMO)

3.2.4

Outgoing Returns Volume and Value Data

The following table illustrates the estimated annual volume and dollar value of returned checks
and the respective 95% confidence interval for each estimate.
Table 13: Estimated Annual Aggregate Outgoing Returns (Volume and Value)
Category

Estimate

95% Confidence Interval

% Total Checks

Returns Volume

251million

(+/-)

20.6million

0.59%

Returns Value

176billion

(+/-)

15.9billion

0.45%

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

Chapter 4

Check Sample Study

The Check Sample Study (CSS)

In an effort to characterize the check payments market, the survey research team conducted a
two-stage stratified random sample survey of bank-of-first-deposit (BFD) checks.27 The
characteristics of each check were recorded on the survey form and each check then categorized
in terms of payer, payee and purpose of the transaction.
Privacy was a great priority of The Check Sample Study (CSS). No sensitive information or
characteristics of the check that could identify the payer, payer account, payee or payee account
were collected. To further ensure the privacy of the parties to each check, access to the sampled
checks was limited strictly to the participant financial institutions (or their processors), who
collected all data themselves. Neither Global Concepts, Westat nor the Federal Reserve collected
any data directly from the sampled checks. The survey instrument allowed respondents to
identify, for example, whether particular phrases or suffixes, such as "Inc., LLC, PLC, Corp,
LTD or .com" were present on a sampled check without having to record the actual name of the
payer or payee.
Data were collected retrospectively for the reference period May, 2000 to April, 2001. By
overlapping the reference period with that of The DFI Check Study – March and April 2001 – the
methodology ensures maximum compatibility of the two studies' results.
4.1

Methodology

4.1.1

CSS Sample Design

The Check Sample Study required a two-stage sample design in order to survey a representative
random sample of checks:
•

For the first stage of the sampling for this study, an approach similar to The DFI Check
Study was used, with the same three institution type strata constructed, followed by sizebased (i.e., PCD-based) strata within each. For commercial banks, a certainty stratum was
established that contained many of the same banks in the first certainty stratum of The
DFI Check Study.

•

The second stage of the sampling addressed the allocation of checks. Since the goal of the
CSS was to describe the universe of checks, the sample for the CSS was designed to

Because The Check Sample Study sampled BFD items, it includes all check volume, including Treasury checks,
postal money orders and travelers checks.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

34
ensure the same probability of selection for each check in the U.S. that was deposited
during the reference period.
Although this was a study of checks, minimizing the number of institutions to be recruited to
provide checks made the data collection effort more efficient and cost effective. Additional
reasons for using this approach for this stage were that it was similar to The DFI Check Study
approach, enabling comparison to The DFI Check Study findings, and it ensured representation
of the largest institutions, especially the top 100 commercial banks.
4.1.1.1 Sample Size and Sample Allocation

The number of sampled institutions and strata for each type are reported in the table below.
Table 14: Stage One Sample Allocation – DFIs Sampled per Stratum
Stratum
Top '100'
1
2
3
4
5
TOTAL

Commercial Banks
87
31
42
31
59
62
312

Type of Institution
Credit Unions

Thrifts

41
31
23
40

17
43
44

135

104

TOTAL
87
89
116
98
99
62
551

Commercial Banks
12,451
1,523
1,952
1,475
2,808
2,951
23,160

Type of Institution
Credit Unions
2,080
1,475
1,095
1,904
6,554

Thrifts
2,146
2,046
2,094
6,286

TOTAL
12,451
5,749
5,473
4,664
4,712
2,951
36,000

The overall approach of the design was to achieve nearly the same probability of selection for
each check in the sample universe. Selected institutions were asked to report their overall
deposit volume as part of the preliminary information for the CSS. Institutions could then have

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

35
their responses weighted by deposit volume adjusted for the number of sampled checks and PCD
value.
4.1.2

CSS Sample Weighting

Data collected from The Check Sample Study (CSS) were inflated to the universe level using
weights designed to (1) compensate for unequal selection probabilities; and (2) adjust for nonresponding depository financial institutions (DFIs). As discussed above, the CSS employed a
two-stage sample design to select the sample of checks. The DFIs were selected at the first stage
of sampling, and the checks were sampled in the field at the second stage from the sampled
DFIs. The DFIs in the certainty strata were all selected for the study. The sample of DFIs from
the non-certainty strata was selected with probability proportional to size (PPS) sampling from
within strata. The sample of checks was selected independently from each sampled (respondent)
DFI using systematic sampling procedure. Two weights were constructed corresponding to the
two stages of sampling, i.e., a DFI weight to represent the non-sampled and non-respondent DFIs
and a check weight to represent the non-sampled checks within the respondent DFIs. The final
CSS weight was obtained by multiplying the two weights, i.e., the DFI and check weights.
The base weight for a sampled DFI was defined to be the reciprocal of selection probability of
the DFI. The base weights were adjusted to account for the non-responding DFIs. The
nonresponse adjustment factor was applied within each stratum. The nonresponse adjustment
factor was defined as the ratio of the number of DFIs sampled from the stratum and the number
responding from that stratum.
This weight adjustment was applied to increase the weights of the sampled DFIs for which data
were collected. As discussed above, the nonresponse weight adjustment was applied at the
stratum level. The weighted adjusted for nonresponse were simply the product of the base
weights and the nonresponse adjustment factor for the stratum.
4.1.3

Estimation

Based on the sampling and weighting methods described above, the study's results allowed us to
estimate the proportion of checks that satisfied each of a series of classification criteria as
discussed in section 4.1.8.
4.1.4

Random Sampling in the Field

The objective to achieve randomness of the sample within each DFI presented a unique
challenge for the survey research team. It was important that checks not all be sampled from the
same date, obviously, but also not from the same processing facility, sorter device, time of day,
roll of microfilm, etc. A bank with a large corporate customer, who always deposits a large
volume of checks early in the morning, for example, could seriously bias that bank's sample if all
checks were sampled from the first roll of microfilm for each of the randomly selected days. The
sampling process required that many variables be randomized to ensure the most representative
random sample.
4.1.4.1 The Sampling Parameters Request Form

In an effort to maintain as much methodological control as possible over the sampling process,
while at the same time sampling in the most appropriate and efficient way for each institution,

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

36
the survey research team developed the Sampling Parameters Request Form (Appendix E). The
form was a pre-survey instrument or screener that allowed each DFI to describe the environment
in which checks would be sampled. An institution could indicate, for example, its number of
processing facilities; the average monthly volume of checks captured at each facility; whether its
check archival system assigns a unique trace number or sequence number to each item in
archive; and whether checks are archived on microfilm, digital image media or a combination of
the two.
This information allowed the survey research team to design a customized set of sampling
instructions for each DFI that completed the Sampling Parameters Request Form – instructions
that ensured the most random and representative sample of checks. If an institution had multiple
processing facilities, for example, the survey research team specified exactly how many checks
should be sampled from each of the institution's processing facilities, from what dates and from
exactly where in the sequence of each day's check processing volume.28
4.1.4.2 Retrieval of Randomly Selected Checks

The survey research team provided each participant institution with a Master List of Random
Checks, which accompanied each institution's sampling instructions. The Master List included
for each item to be sampled the date on which it was processed and a specific Random Check
Number. Dates were chosen at random from the 252 eligible processing days in the May 1, 2000
to April 30, 2001 survey period. The Random Check Number for a given date was chosen at
random between 1 and the institution's average daily volume of checks processed.
The average daily volume was calculated using volume data provided via the screener. For
institutions that commingle deposited checks with inclearings, the upper bound on this Random
Check Number was the average daily prime pass volume – essentially, the combined volume of
both inclearings and deposits. This ensured that each deposited check had an equal probability of
being selected from the commingled archive, regardless of whether it fell at the end of a batch of
inclearings. For institutions that archive deposited checks separately from inclearings, the upper
bound on the Random Check Number was simply the average daily deposit volume.
As a practical matter the notion of a Random Check Number worked well for smaller institutions
but was overly simplistic for many of the larger institutions in the survey. The primary method of
randomizing the sample for many institutions was through the use of Random Sequence
Numbers. Global Concepts worked closely with dozens of institutions to customize Master Lists
of Random Sequence Numbers to facilitate random selection of sample checks.
4.1.4.3 Photo Retrieval Latitude

There was no guarantee, unfortunately, that a Random Check Number or Random Sequence
Number provided by Global Concepts would point to an actual deposited check. In some cases,
the Random Number would exceed the total volume processed that day. In other cases, the
Master List may have listed an item for which the DFI was not the bank of first deposit, a deposit

It is common for a financial institution to have a single archive for checks processed at multiple capture facilities.
All checks were, of course, sampled from the central archive in these situations, but they were sampled in such a
way as to accurately represent the distribution of check volume across multiple capture facilities.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

37
slip or a general ledger (GL) ticket.29 For that reason each Master List included alternate Random
Check Numbers or Sequence Numbers to increase the likelihood of a "hit" on a given date. Even
these were no guarantee. Therefore, all institutions were instructed to find the deposited check
nearest to the Random Check Number or Sequence Number specified. Generally, this meant
rewinding microfilm or advancing to the next item in an optical image archive.
4.1.5

Designing the Survey Instrument

In designing the survey methodology, Global Concepts and the Federal Reserve decided that a
survey of deposited checks (i.e., BFD items) was the most appropriate way to survey a
representative random sample of check payments.30
A primary motivation to survey BFD items rather than paid check volume was that BFD items
are simply easier to distinguish from correspondent volume – checks processed (and archived)
on behalf of another institution. Bank of first deposit items, unlike other transactions, have only
one financial institution endorsement – the survey respondent's. This seemed to the research
team like a more practical way of distinguishing a bank's own volume from another institution's
than by reviewing both the endorsement on the back of the check and the RT and/or payor bank
name on the front of the check.
A survey of BFD items also ensured that checks to or from the federal government would be
sampled along with private sector checks.
4.1.5.1 Pretesting the Survey Instrument

The Check Sample Study survey instruments and sampling instructions were pretested
concurrently with The DFI Check Study survey instruments. Pretest participants' support was
invaluable in creating the final survey instrument.
4.1.6

The Survey Instrument

A copy of The Check Sample Study survey instrument can be found in Appendix D. In the field,
a more lengthy set of instructions accompanied the survey instrument. The instructions included
visual aids and helpful hints to assist in the identification and recording of survey data.
4.1.7

Reporting Period: May 1, 2000 - April 30, 2001

The survey research team decided that a 12-month survey period was the best way to account for
seasonal variation in the distribution of checks written in the U.S. Unlike The DFI Check Study,
The Check Sample Study could not rely on Federal Reserve seasonality data as an adjustment
factor to a 2-month survey. The Federal Reserve data address only aggregate volume variation –
not variation in descriptive data, such as the purpose, payer or payee of check payments.
The choice of a 12-month survey tackled the seasonality problem, and it proved uniquely
practical from a data collection perspective. A retrospective 12-month survey posed no difficulty
for respondent DFIs, which archive check deposits for up to 7 years. Unlike the survey of check
29

BFD items are distinguished by the presence of only one DFI endorsement.
No strong evidence existed to suggest that a survey of deposited checks would produce less cluster distortion than
a survey of paid checks. We assumed that by surveying either BFD items or paid checks we faced roughly equal risk
of cluster distortions based on the homogeneity of customers at individual respondent DFIs.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

38
volume (for most institutions), The Check Sample Study required no advance notice in order to
collect the data in a timely manner. Check archival practices provide a unique environment in
which to conduct retrospective rather than prospective survey research.
As mentioned previously, the reference period of May 1, 2000 to April 30, 2001 also allowed for
maximum compatibility of results between this study and The DFI Check Study, which surveyed
check volume between March 1, 2001 to April 30, 2001.
4.1.8

Check Payments Categorization

In designing the methodology for The Retail Payments Research Project, Global Concepts and
the Federal Reserve determined that documenting the opportunity for substitution of electronics
for paper transactions required additional descriptive characteristics beyond the total volume and
dollar value of payments. We determined three factors to be sufficiently measurable and
descriptive to document substitutability: the type of payer, type of payee and purpose of the
transaction. Each of these factors was in turn subdivided into its respective categorization options
below:
Table 16: Payments Classification Factors
Payer

Payee

Purpose

Consumer

Remittance

Business

Point of Sale (POS)

Government

Income Payments31
Casual Payments32

4.1.8.1 Payer and Payee Categories

During the design phase of The Retail Payments Research Project, Global Concepts and the
Federal Reserve decided that three categories – Consumer, Business and Government –
sufficiently described the potential parties to a payment.
•
•
•

Consumer – an individual, household or small business33
Business – a private sector entity
Government – local, state or Federal government entity

These categories are not only commonly accepted in the industry, but they make an appropriate
delineation between the types of electronic payment alternatives that may be available to or
accepted by the respective parties to any payment.
31

Income describes any payment from a business or government entity to a consumer (i.e., individual) or small
business indistinguishable from a consumer.
32
Casual describes any payment from one consumer (i.e., individual) to another. This also includes small businesses
that are indistinguishable from consumers.
33
Some small business owners (e.g., sole proprietorships) use their personal checking accounts for business
purposes and cannot be distinguished from consumers based on data from their checks alone.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

A consumer, for example, generally does not have the ability to accept credit or debit card
payments; whereas a business or government would have no significant impediments to
accepting debit or credit as alternatives to paper checks at the point of sale. As a payer, a
consumer is likely to have online debit available to him or her; whereas, a business or
government entity cannot pay with online debit.
Very small businesses, such as sole proprietorships, may resemble a consumer payer or payee
more closely than a business in terms of availability and use of electronic payment alternatives.
As a practical matter, The Check Sample Study effectively deals with the commonality between
consumers and sole proprietorships by assuming that any check written to or from an individual
and having no characteristics on the check to indicate a business payer or payee is classified as
consumer payer or payee respectively.
The distinction between business and government is largely immaterial for the purpose of
evaluating substitution potential. There are no particular impediments to a government entity
accepting a payment type that a business might accept and vice versa. Likewise, business or
government payers are expected to have comparable access to the same payment mechanisms,
such as purchasing cards, financial EDI or ACH initiation capabilities.
4.1.8.2 Purpose Categories

Considering all possible payment types and their various options for substitution of electronic for
paper payments, Global Concepts and the Federal Reserve determined that all payments fall into
one of four primary purpose categories:
•
•

•

Casual – Payment from one individual to another. By definition, all consumer-toconsumer payments, therefore, are categorized as Casual. As a category, these payments
are believed to have a relatively low potential for electronic substitution.
Income – Payment to an individual from either a business or government entity. By
definition all business-to-consumer or government-to-consumer payments, therefore, are
categorized as Income. This category includes:
o Payroll
o Pension
o Benefits / Entitlements
o Rebate / Promotional / Refund
o Expense Reimbursement
o Tax Refunds
o Investment Disbursements
o Remittances to Small Businesses Indistinguishable from Consumers
Remittance –Payments from any type of payer to either a business or government payee
that does not occur at the point of sale. The types of remittance payments include:
o Recurring Retail Remittance – Regular recurring payments, typically described as
“bill payments.” Examples: utility bill payments, insurance premiums, telecom
charges, credit card bill payments, loan repayments, etc.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

•

o Non-Recurring Retail Remittance – Irregular remittance payments made for
products or services rendered for consumer consumption. Examples: medical bill,
plumber, carpenter, pest control, legal fees, accountant fees, etc.
o Commercial Remittance – Any B2B, B2G, G2B, G2G payments not made at the
Point-of-Sale. Examples: raw materials purchase, office supplies, business
equipment, finished goods from wholesalers, etc.
Point of Sale (POS) – Payments from any type of payer to either a business or
government payee that occurs in any of the following environments:
o Storefront – Traditional single or multi-lane retail environment, such as
department store, drugstore, clothing store, gas station, dry cleaner, concessions,
etc.
o Over-The-Counter Retail Remittance – Remittance payments made in person,
such as telecom bill paid at the local office, utility bills paid in person, medical
expenses paid at the doctor’s office, etc.
o MOTO – Mail Order/Telephone Order transactions (e.g., catalog orders).
o Internet – Purchase of goods or services over the Internet.
o Mobile POS / C.O.D. – Payments made for goods or services delivered offpremise by the seller with payment occurring at time of delivery, such as food
delivery, home maintenance fees, etc.
o Vending.34

The intersection of the three payer types, three payee types and four purpose classifications visà-vis check payments is described in the matrix that follows. Note that gray shaded cells indicate
check payment types that do not exist.35

Internet, MOTO, and Vending transactions fall into the POS category, but they do not apply to check payments.
The categorization was designed to describe all potential payment mechanisms – not just Check – for the POS
purpose category.
35
It was decided that dividend payments to corporate shareholders would not qualify as Income payments. From a
substitution perspective – i.e., the ability to substitute electronic for paper payments – this category is
indistinguishable from business-to-business remittance payments and, therefore, should be categorized as such.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

41
Exhibit 8: Original Check Categorization Matrix
P U R P O S E

P A Y E E

P A Y E R
C

C
REMITTANCE

T
C = Consumer

B = Business

G
T

G = Government
T = Total

C
POINT-OF-SALE

B
G
T
C

INCOME PAYMENTS

B
G
T
C

CASUAL PAYMENTS

B
G
T
C

TOTALS (VALUE, VOLUME)

B
G
T

A primary purpose of The Check Sample Study was to document the distribution of a 12-month
check payments market across this matrix.
4.1.8.3 Groupings of Categories for Data Analysis

For the purpose of reporting data and performing data analysis, a number of categories were
grouped into single cells. This helped to simplify the analysis and also to create more meaningful
cell sizes for analysis. The groupings used in our analysis are as follows:
4.1.8.3.1 Payer and Payee Groups

Whether payer or payee, the categories business (B), government (G) and business or
government (BG) have been grouped into a single business or government category. From a
substitution perspective, business or government entities are indistinguishable. It should be noted
that the vast majority of business or government checks (whether payer or payee) are business
checks.
The business or consumer (BC) category allows us to measure the extent to which business and
consumer names (and checks) are difficult to differentiate. From a substitution perspective,
however, the two have little in common. Consumers do not accept POS or remittance payments;
nor do businesses use debit cards. Therefore, the BC category (whether payer or payee) was
combined with indeterminate and "Error" categorizations into an unknown category for analyses.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

42
Payer / Payee
B, G, or BG
BC, Indeterminate (X), Error

Grouped as
Business or Govt
Unknown

4.1.8.3.2 Purpose Groups

Only checks of indeterminate purpose and those for which the purpose categorization is believed
to be in error were combined in the analyses of purpose data.
Purpose
Indeterminate (X), Error
4.1.9

Grouped as
Unknown

The Check Categorization Model

Global Concepts employed a categorization model based on conditional logic to assign a
classification to each check. Judging from the characteristics of the check, as described by the
DFI respondent, the model assigned a payer, payee and purpose classification to each item. If the
model could not definitively categorize the surveyed item, it was categorized as Unknown.36
The types of factors that went into the categorization of each item are described below.
4.1.9.1 Categorization of the Payer

The determination of the payer of the check was made entirely from information available on the
face of the check.
Checks categorized as Business, Government or Business or Government were typically
identified by their larger format, the characteristics of the MICR line (e.g., Federal Government
checks' MICR line begins with 000, many business checks include unique MICR characters), the
method used to frank the check (e.g., typed or machine printed "signature"), and the
characteristics of the payer name and address. The payer name/address was particularly useful in
identifying the presence of such indicators as Inc., LLC, PLC, LTD, Corp., Department of, City
of, Town of, Bureau of, Accounts Payable, etc. It was also extremely useful to individual
respondents, who were asked to conclude what type of payer had written the check and to
indicate why this classification was selected. The payee line (e.g., following "Pay to the order
of…") was also useful in some cases, because business or government payers – unlike consumers
– sometimes include the full mailing address of the payee (machine printed) on the face of the
check.

In practice Unknown could result from one of three outcomes: Business or Consumer, Indeterminate or Error.
The model returned a Business or Consumer outcome if no conclusion could be reached about the check other than
to eliminate Government as a potential classification. Since Business or Consumer provides no useful information
about the check in terms of its potential likelihood to be substituted by an electronic payment in the future, Unknown
was the most appropriate conclusion. The model returned an Indeterminate outcome if the survey form was correctly
completed but the data were still inconclusive. An Error outcome resulted if the survey form contained
contradictory data.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

43
Checks classified as Consumer generally included smaller format checks without characteristics
in the MICR line or name/address fields to suggest a business or government classification.
Respondents were also asked whether they believed the check to be a consumer check and why.
Note: It is entirely possible that some small businesses or sole proprietors might use their
personal checks for business payments. Without any characteristics to indicate a business use,
these checks would be classified as Consumer. Considering, however, that a payer of this nature
is for all practical purposes (particularly when considering the substitution possibilities) acting as
a consumer would act, these potential misclassifications are acceptable and, arguably,
appropriate.
4.1.9.2 Categorization of the Payee

The determination of the payee was made from information on both the front and back of the
check: the payee line, the endorsement and any other writing/stamp/print on the check.
Just as they used the payer name/address, respondents were asked to determine whether any
telltale signs of a business or government payee existed in the payee line, e.g., Inc., LLC, Corp.,
IRS, Tax Commissioner, Bureau of, Town of, County of, etc. Additionally, respondents recorded
the presence of unique printing or stamps on the checks that might indicate a POS transaction,
such as the driver's license number, date of birth, such phrases as, store number, terminal
number, cash back, etc. The endorsement was a major determinant of payee type. Business or
government payees tend to stamp or machine print their endorsements on the back of checks.
Lockbox (i.e., remittance) payments in particular tend to be endorsed along the length of the
check (i.e., parallel to text on the face of the check) rather than across the end of the check (i.e.,
perpendicular to text on the face of the check). These indicate business or government payees.
In all cases, the respondent was also asked to indicate his/her determination of the payee of the
check based on all the information available (payee line, endorsement, etc.) and to indicate why
this determination was made.
The payee classification of Consumer was made if a) the check showed no indications of being
written to a business or government payee and b) this fact agreed with the respondent's
determination and explanation that the payee was a consumer.
4.1.9.3 Categorization of the Purpose

The purpose of the check payment was determined by a combination of information on the check
itself and the classification of its counterparty (i.e., payer and payee).
The payer and payee relationship (counterparty) alone was enough to determine the purposes of
some checks. For example, all business-to-consumer, government-to-consumer, or
business/government-to-consumer checks were classified as Income. Note: Not all income
payments as categorized by this study are payroll checks. Rebate checks, tax refunds, stock
dividends are all types of checks that would fall in the Income category. Similarly, all checks
payments from one individual to another individual were classified as Casual. Based on the
examples discussed above, this category no doubt includes payments to or from sole
proprietorships or small businesses that use what are, or appear to be, personal checks for
business transactions. Casual might also include payments from an individual to his/her attorney,

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

44
and rent payments from tenants to landlords are almost certainly included in Casual. While the
term "casual" may incorrectly define the intent of many of these transactions, the classification
Casual is not entirely inappropriate. In many cases these checks have a low probability of
substitution by electronic payment. It is appropriate to group these checks with payments
between two individuals. We acknowledge the potential for these misclassifications but believe
they are acceptable for the purposes of this study.
Any check written to a business or government payee was categorized as either Remittance or
POS. If the distinction could not be made, these checks were categorized as Remittance/POS.
The distinction was made by using information about the type of organization being paid and by
the characteristics of the endorsement. If the payee was clearly a credit card issuer, a utility, etc.
this lent evidence toward a Remittance classification. Conversely payments made to a
convenience store, a restaurant, or a drugstore would suggest the payment was made at the point
of sale. This information was evaluated alongside information about the endorsement. If the
endorsement included such information as a store number, a terminal number or a customer's
driver's license number, this suggested a POS transaction. Lockbox endorsements (typically
apparent by their alignment across the length of the check) or the terms "absentee" or "absent
endorsed" indicated a remittance payment.
4.1.10 Data Review

Global Concepts performed data review and validation on the survey responses both to evaluate
the accuracy and completeness of the survey responses and to test the function of the
categorization model. Neither exhibited any systematic problems.
4.2
4.2.1

Results and Analysis
Survey Response

In total 27% of financial institutions completed the Check Sample Study. The table below
illustrates the response rate by type of institution and size stratum.
Table 17: CSS Response Rate by Stratum
Stratum
Top '100'
1
2
3
4
5
TOTAL

Commercial Banks
67%
16%
26%
13%
20%
18%
32%

Type of Institution
Credit Unions

Thrifts

27%
19%
13%
20%

35%
21%
14%

21%

20%

TOTAL
67%
25%
22%
13%
20%
18%
27%

Respondent DFIs sampled and surveyed 28,877 checks in total. The distribution of those checks
across each type of institution and size stratum can be found in the following table.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

45
Table 18: Checks Surveyed for CSS by Stratum (Actual Response)
Stratum
Top '100'
1
2
3
4
5
TOTAL

Commercial Banks
19,145
450
980
359
1,117
988
23,039

Type of Institution
Credit Unions

Thrifts

979
499
270
696

2,068
807
519

2,444

3,394

TOTAL
19,145
3,497
2,286
1,148
1,813
988
28,877

The largest banks not only provided the greatest number of checks, but they did so by responding
at a higher than expected rate. The table below illustrates the checks received by each stratum as
a percentage of checks expected according to the design of the survey. Note: Weighting
adjustments applied to the sample help to offset bias introduced by the higher than expected
response rate among large commercial banks.
Table 19: Response Rate (in Checks) as a Percentage of the Original Design
Stratum
Top '100'
1
2
3
4
5
TOTAL

4.2.2

Commercial Banks
154%
30%
50%
24%
40%
33%
99%

Type of Institution
Credit Unions

Thrifts

47%
34%
25%
37%

96%
39%
25%

37%

55%

TOTAL
154%
61%
42%
25%
38%
33%
80%

Survey Results

The Check Sample Study distributes the volume and value of check payments in the U.S.
according to payer, payee and purpose. The following tables detail the study's results.
4.2.2.1 Distribution of Check Volume

The tables in this section detail the distribution of check payments volume according to payer,
payee, counterparty and purpose by counterparty. Each sub-total data element in the tables below
includes a corresponding error estimate.37 The error is the 95% confidence interval around each
estimate. Consumer-to-consumer casual payments, for example, represent 11.2% of all check
payments plus or minus 1.9%; or 9.3% to 13.1% of all check payments.

Error estimates are not provided for point estimates of 0.1% or less.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

46
Table 20: Estimated Distribution of Check Volume by Payer
Check Volume by Payer with 95% confidence interval

Consumer
Business
Government
Business or Government
Unknown
Total

% total
50.9 %
32.3 %
3.5 %
1.3 %
12.0 %
100.0 %

(+/-)
2.2 %
2.1 %
0.7 %
0.1 %
2.6 %

Table 21: Estimated Distribution of Check Volume by Payee
Check Volume by Payee with 95% confidence interval
% total
29.1%
50.1%
2.7%
3.8%
14.3%
100.0 %

Consumer
Business
Government
Business or Government
Unknown
Total

(+/-)
4.4 %
3.2 %
0.5 %
0.6 %
2.5 %

Table 22: Estimated Distribution of Check Volume by Counterparty
Distribution Includes 95% confidence interval
Payer
Consumer
Business
Government
Bus or Gov
Unknown
Total

Cons
11.2%
14.6%
2.5%
0.6%
0.1%
29.1%

(+/-) Bus
1.9% 33.8%
1.9% 15.0%
0.6%
0.7%
0.1%
0.4%
0.2%
50.1%

RETAIL PAYMENTS RESEARCH PROJECT

(+/-)
3.0%
1.6%
0.2%
0.1%
0.1%

Payee
Govt (+/-) Bus/Govt
1.7% 0.3%
2.7%
0.8% 0.2%
1.0%
0.2% 0.1%
0.0%
0.0%
0.1%
0.0%
0.0%
2.7%
3.8%

(+/-) Unknown
0.4%
1.6%
0.2%
0.9%
0.1%
0.1%
11.6%
14.3%

(+/)
0.3%
0.2%

2.6%

Total
50.9%
32.3%
3.5%
1.3%
12.0%
100.0%

Sponsored by the Federal Reserve System

47
Table 23: Estimated Distribution of Check Volume by Counterparty and Purpose
Distribution Includes 95% confidence interval

Purpose
Income
Casual
Remittance
POS
Remit/POS
Unknown
Total

C2C

(+/-)

C2BG

(+/-)

Counterparty
BG2C (+/-) BG2BG
17.8% 2.5%

(+/-)

Unknown

(+/-)

0.7%
0.8%
0.9%

0.1%
0.1%
0.1%
14.5%
14.7%

0.1%
0.0%
0.0%
2.5%

11.2% 1.9%
17.7% 1.9%
14.1% 2.0%
6.4% 0.9%
11.2%

38.1%

7.9%
4.9%
5.5%
17.8%

18.3%

Total
17.8%
11.2%
25.7%
19.0%
11.9%
14.5%
100.0%

4.2.2.2 Distribution of Check Value

The tables in this section follow the same conventions as section 4.2.2.1, but refer to the dollar
value rather than the volume of check payments. As in the previous section, error estimates are
excluded for point estimates of 0.1% or less.
Table 24: Estimated Distribution of Check Value by Payer
Distribution Includes 95% confidence interval

Consumer
Business
Government
Business or Government
Unknown
Total

% total
19.2%
61.6%
3.8%
1.5%
13.9%
100.0 %

(+/-)
2.1%
3.5%
1.1%
0.5%
2.9%

Table 25: Estimated Distribution of Check Value by Payee
Distribution Includes 95% confidence interval

Consumer
Business
Government
Business or Government
Unknown
Total

RETAIL PAYMENTS RESEARCH PROJECT

% total
23.3%
55.5%
1.5%
3.8%
15.9%
100.0 %

(+/-)
4.0%
4.3%
0.6%
1.9%
2.9%

Sponsored by the Federal Reserve System

48
Table 26: Estimated Distribution of Check Value by Counterparty
Distribution Includes 95% confidence interval
Payer
Consumer
Business
Government
Bus or Gov
Unknown
Total

Cons
6.7%
14.5%
1.5%
0.4%
0.1%

(+/-) Bus
1.2% 10.2%
3.2% 42.7%
0.5%
1.9%
0.1%
0.5%
0.2%

23.3%

55.5%

(+/-)
1.8%
4.6%
1.0%
0.3%
0.1%

Payee
Govt (+/-) Bus/Govt
0.6% 0.4%
0.7%
0.5% 0.3%
2.9%
0.3% 0.2%
0.0%
0.0%
0.1%
0.0%
0.0%
1.5%

(+/-) Unknown
0.2%
0.9%
1.8%
0.9%
0.0%
0.5%
13.6%

3.8%

(+/)
0.4%
0.3%

Total
19.2%
61.6%
3.8%
1.5%
13.9%
100.0%

2.9%

15.9%

Table 27: Estimated Distribution of Check Value by Counterparty and Purpose
Distribution Includes 95% confidence interval

Purpose
Income
Casual
Remittance
POS
Remit/POS
Unknown
Total

C2C

(+/-)

6.7%

1.2%

6.7%

C2BG

(+/-)

Counterparty
BG2C
(+/-) BG2BG
16.5% 3.6%

6.2%
1.7%
3.7%

1.5%
0.4%
0.9%

21.1%
7.1%
20.8%

11.5%

16.5%

(+/-) Unknown

4.9%
2.2%
4.6%

49.0%

0.1%
0.0%
0.1%
16.1%
16.2%

(+/-)

0.1%
0.0%
0.1%
2.9%

Total
16.5%
6.7%
27.4%
8.8%
24.6%
16.1%
100.0%

4.2.2.3 Average Dollar Value by Payment Purpose and Counterparty

The table below details the average dollar value per check in each of twelve categories. The
corresponding 95% confidence interval (in dollars) is provided alongside the average dollar
value figure for each category.
Table 28: Estimated Average Value per Check Category (Counterparty by Purpose)
Distribution Includes 95% confidence interval
Counterparty
C2C
(+/-)
C2BG
(+/-)
BG2C
(+/-)
BG2BG
(+/-) Unknown
Purpose
Income
$1,018.65 $126.77
Casual
$663.47 $134.88
Remittance
$ 387.52 $ 51.99
$2,914.46 $667.74 $ 763.39
POS
$ 128.40 $ 21.56
$1,614.68 $484.48 $ 288.38
Remit/POS
$ 629.91 $158.47
$4,169.19 $814.01 $1,366.46
Unknown
$1,220.00

RETAIL PAYMENTS RESEARCH PROJECT

(+/-)

$ 632.97
$ 184.19
$1,639.61
$ 286.16

Sponsored by the Federal Reserve System

49
4.2.2.4 Distribution of Checks by Dollar Value Category

Based on industry feedback during the preparation of this report, the Federal Reserve decided to
include data about the distribution of check payments across dollar value ranges. Given the late
inclusion of these data, no error estimates have been computed. While we have not computed
error estimates, the percentage values are comparable to other values reported in this report; we
would expect their error estimates to be comparable as well.
The majority of checks appear to be written for relatively low dollar transactions. As illustrated
in the table below, nearly a third of all checks (32%) are written for $50 or less. Over 75% of all
checks are for transactions of $500 or less.
Table 29: Distribution of Checks by Dollar Amount
Dollar Amount Range
$0.01-$50
$50.01-$100
$100.01-$500
$500.01-$1,000
$1,000.01-$2,500
$2,500.01-$5,000
$5,000.01 +
Total

RETAIL PAYMENTS RESEARCH PROJECT

% Check Payments
33.3 %
14.7 %
28.6 %
10.1 %
6.3 %
3.5 %
3.4 %
100.0%

Sponsored by the Federal Reserve System

Chapter 5
Electronic Payment
Instruments Study

Electronic Payment Instruments Study

5.1

Introduction

5.1.1

Objectives

The primary purpose of this research was to determine the volume and value of retail electronic
payment transactions originating in the United States for the year ending December 31, 2000.
This information - including information on consumer, business and government initiated
electronic payments and remittances - will provide valuable input into the policy and longer-term
operational decision-making of the Federal Reserve Bank. The survey captured data on the
following electronic payment instruments:
•
•
•
•

General-purpose and private-label credit cards
Online and offline debit cards
Automated Clearing House (ACH) transactions
Electronic Benefits Transfer (EBT) payments

In addition, information on Emerging Payment Instruments (including open system stored value
and Internet payments) was tracked for informational purposes.
Participation in the study was voluntary but was encouraged by the Federal Reserve team
through industry wide communications and personalized letters.
5.1.2

Scope

The Electronic Payment Instruments Study (EPIS) collected data on electronic payments made in
the U.S. during the year 2000. Transactions from consumers, businesses and government entities
have been included in the statistics gathered. Data were gathered in three primary areas:
1. Electronic payment options used by buyers of goods or services, including point-of-sale
transactions.
2. Electronic payment products used on the ‘back-end’ to effect final settlement for
purchase transactions, including bill payment.
3. Electronic payment options used by employers, state agencies and others for
disbursements of income payments such as payroll and benefit disbursement transactions.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

51
The types of transactions included in the study and organizations surveyed are summarized in the
following table:
Table 30: Types of Transactions and Organizations Included in Study
Transaction Type
General-purpose credit/charge, offline debit
Private-label credit/charge
Online debit
Automated Clearing House (ACH)
Electronic Benefits Transfer (EBT)
Emerging Payments

Organization Type Surveyed
Credit and charge card associations
Retailers, oil companies, fleet card issuers,
processors, third party receivables owners
Regional and national EFT networks
NACHA, ACH operators
USDA Food and Nutrition Service, EBT
contractors
Companies involved in bill payment, P2P,
stored value, Internet currencies, and other
new payment technologies

Organizations involved in emerging payments were surveyed for informational purposes only.
Most of these new payment types are a new front-end payment method to the consumer, but use
traditional funding and settlement systems behind the scenes. Adding their volume numbers into
the aggregate would result in double-counting.
5.1.2.1 Excluded from Scope

Only unique payment instruments and their final settlement were tracked for this study. This
excludes paper-based transactions (i.e., cash, checks) and exchanges within a payment
instrument where no value has changed hands.
Additionally, there are variations of payment instruments as well as components of the payments
value chain that the Federal Reserve considered to be outside the scope of this study:
•
•
•
•

•
•
•
•
•

Cash and check transactions
Electronic bill presentment transactions
Bill payment transactions which are paid via paper (even if initiated electronically)
Closed system stored value purchases, including:
o Gift cards
o Internet currencies
o Loyalty-based accounts (e.g., airline frequent flier accounts)
o Phone cards
o University and military closed payment systems
On-us online debit transactions
On-us ACH transactions outside of the estimates provided by NACHA
Cash Concentration through ACH
Consumer and business wire transfers
Issuer-to-acquirer settlement transactions

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

52
5.2

Research Methodology

The study was conducted by surveying the leading electronic payment processors. The primary
sources for this information were major card industry associations and processors, Federal
Government agencies, EFT networks and other entities that could provide accurate and reliable
data. The survey was conducted during the spring of the year 2001 and surveys were sent out to
178 organizations within 157 companies that either originate or monitor payment transactions in
the United States.
5.2.1

Participation Rates

Overall response to the study exceeded expectations: 75% of electronic payment companies
participated in the survey, providing data on 94% of all estimated electronic payment value in
2000. Transaction and dollar values for organizations not participating in the study were
estimated.
By payment instrument, participation rate of debit providers was highest and credit card
participants was lowest. However, all payment instruments had very strong participation in the
study.
Table 31: Study Participation Rate by Payment Instrument
Payment Instrument

Potential
Participants

General-purpose Credit Cards

Participation Rate by
Organization

Transaction
Volume

Dollar Value

71%

76%

73%

69%

49%

44%

Offline Debit

100%

Online Debit

85%

95%

88%

ACH

100%

EBT

80%

88%

87%

118

75%

84%

94%

30%

Private-Label Credit Card

Total
Memo: Emerging Payments

The high level of participation by the largest payment industry companies provides confidence
that the study results are representative for all payment instruments. The study results provide a
conservative and reliable estimate of electronic payment transactions in 2000.
5.3

Summary of Findings

During calendar year 2000, 29.5 billion electronic payments were originated in the United States
with a value of $7.3 trillion.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

53
Table 32: Total Estimated Volume and Dollar Value of Electronic Payments
Transaction
Volume
(Millions)

Electronic Payment Instrument
General-purpose credit cards

Average
Payment
Value

Dollar
Volume
($Millions)

12,300.2

$1,072,555

$87.20

Private-label credit cards

2,748.6

$162,819

$59.24

Offline debit

5,268.6

$209,980

$39.85

Online debit

3,010.4

$138,151

$45.89

Automated Clearing House (ACH)

5,622.0

$5,674,851

$1,009.40

Electronic Benefits Transfer (EBT)

537.7

$13,744

$25.56

29,487.5

$7,272,100

$246.62

Total

Please note that these volumes exclude non-purchase transactions such as ATM and settlement
transactions, which are outside the scope of this project.
As shown in the charts below, the majority (51%) of electronic payment transactions were made
using credit cards, but 78% of payment dollars were handled through the ACH.
Exhibit 9: Mix of Electronic Payment Instruments
Transaction Volume

Dollar Value

E BT
2%
ACH

ACH

19%

78%

Ge n e ra l p u rp o s e
c re d it c a rd s

E BT

42%

O n lin e d e b it

Ge n e ra l p u rp o s e

10%

c re d it c a rd s
15%

P riv a t e La b e l

O f f lin e d e b it
18%

P riv a t e La b e l

O n lin e d e b it

c re d it c a rd s

5.3.1

c re d it c a rd s
2%

O f f lin e d e b it
3%

General-Purpose and Private-Label Credit Cards

On a transaction volume basis, credit and charge cards were the most common electronic
payment instrument used in the United States during 2000. Fifteen billion transactions were
originated with a value of $1,235 billion. Average transaction size for general-purpose credit
cards was much larger than that of private-label cards: $87.20 vs. $59.24. Credit cards
accounted for 51% of all electronic payment transactions and 17% of the dollar value. Eightytwo percent of credit card transactions and 87% of value came from general-purpose credit cards.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

54
5.3.2

Online and Offline Debit Cards

Following credit cards, debit cards represented the second most common form of electronic
payment, accounting for 8.3 billion transactions and a dollar value of $348 billion in 2000. On
average, each debit transaction was $42, compared with $87 for the average general-purpose
credit card transaction. In 2000, 64% of transactions and 60% of the value was contributed by
offline debit; 36% of transactions and 40% of value was from online debit.
5.3.3

ACH

Although ACH was the third most commonly used electronic payment instrument with 5.6
billion transactions, it dominates on a dollar value basis accounting for 78% of the monetary
value. The average transaction volume was more than 11 times larger than that of generalpurpose credit card transactions ($1,009 vs. $87).
5.3.4

EBT

EBT volume has increased dramatically due to initiatives at the federal level and significant
efforts by state governments to electronify both food stamps and cash assistance payments
during the 1990s. Nevertheless, EBT was the smallest volume payment instrument with .5
billion transactions and $13.7 billion in value. Note that the “EBT” category in this study refers
to consumer payments using EBT. Government disbursements to financial institutions that hold
EBT funds and those institutions’ reimbursements to merchants for EBT sales are included in the
ACH category.
5.3.5

Emerging Payments

In general, emerging payment volumes for the payment instruments studied were quite small in
2000. Organizations participating in the survey reported 76.2 million transactions involving
$12.7 billion. However, these numbers represent only a small portion of the total emerging
payments market. Many organizations did not respond to the survey because they were very new
or they were in a very competitive market and did not want to reveal their data. The Study did
not attempt to estimate the volumes for non-respondents. Several categories within the emerging
payments group will be important to watch in the coming years, especially person-to-person
payments.
5.4
5.4.1

Analysis of Findings and Methodology
General-Purpose Credit Cards

5.4.1.1 Background

Over the past 40 years, credit cards have become an increasingly popular payment method for
consumers and businesses at the point of sale. Cardholders like the convenience and deferred
payment terms offered by the cards; merchants like the “sales lift” they realize when credit card
customers shop in their stores.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

55
There are several types of card products grouped under the credit-card umbrella. All of these
products enable a cardholder to purchase goods or services with a promise to pay at a later date.
Purchases are “charged” to a line of credit or account established for the cardholder.
Table 33: Types of Card Products
Card Product
General-Purpose
Credit Card

Definition/Features
Minimum payment amount due
at the next statement cycle

Examples
Visa
MasterCard

Interest charged on unpaid
balance
T&E CARD

Card designed for use in hotels
and restaurants

Diners’ Club

Charge Card

Full payment required at the next
statement cycle

American Express Green
card

Private-label
Credit Card

Proprietary cards issued by
single merchant

Department Store card
Gasoline card

May or may not allow balances
to revolve

This section includes volume information for General-purpose Credit Cards, T&E Cards, and
Charge Cards – combined under the heading "General-purpose Credit Card." Private-label credit
cards are addressed in the next section.
There are multiple parties involved in each credit card transaction, as shown in the table below:
Table 34: Parties to Typical Credit-Card Transaction
Party

Definition

Cardholder

Consumer or business to whom credit card has been issued

Merchant

Establishment where consumer is shopping

Acquirer

Bank with whom merchant has an account for settlement of
credit card transactions

Issuer

Bank with whom cardholder has the credit card account

Association

Organization owning the credit card brand, switching and
settling transactions between issuer and acquirer, and
establishing routing and sharing rules for the card network

Third-Party
Processors

Provide processing services on behalf of acquirers or issuers

This category includes transactions on the following types of cards:
•
•

General-purpose credit cards
Co-branded credit cards

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

56
•
•
•
•
•
•

Charge cards
Co-branded charge cards
Secured credit cards
T&E cards
Commercial cards, including business, corporate, and purchasing
New payment technologies that route transactions through the card associations’
networks, such as:
o Money sent through the credit card networks by person-to-person (P2P) payment
systems (e.g., PayPal, Billpoint, or eMoneyMail)
o Amounts charged to a credit card where the original payment mechanism was a
transponder (e.g., Mobil Speedpass)
o Open system stored value cards that route their transactions through the credit
card networks (e.g., Visa Buxx card)

(Information on offline debit transactions routed through the Visa and MasterCard networks is
provided in a later section.)
5.4.1.2 Organizations

Since every credit card transaction must be routed through the card association owning the
brand, card associations were the survey focus for gathering credit card and charge card
transaction and sales volume information. Individual issuers, acquirers and merchants were not
surveyed.
5.4.1.3 Survey Data: General-Purpose Credit Cards

General-purpose credit card associations were surveyed for totals of U.S. originated approved
and settled transactions for calendar year 2000.
The following information was requested of survey respondents:
•
•

Required data:
o Number and value of transactions by type of card used and payment mechanism:
credit, charge, offline debit, online/offline and stored value
Optional data:
o Number and value of transactions by venue: in-store, mail/telephone order,
Internet, other card not present purchases, other card present purchases
o Number and value of transaction by type of retailer
o Number of cards outstanding by type of card: credit, charge, offline debit,
online/offline and stored value
o Number of financial institutions issuing cards
o Number of merchants, merchant locations and card terminals accepting card

(Information on offline debit transactions routed through the Visa and MasterCard networks is
provided in a later section.)

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

57
Transactions are routed through one and only one card association, so there is no risk of double
counting among the seven organizations surveyed. Two specific opportunities for overcounting
or undercounting were addressed:
•

Co-branding
A co-branded card is one that has both the credit card association’s brand on it as well
as that of a retailer or other marketer. In some cases, retailers are replacing their privatelabel cards with a co-branded general-purpose card. These cards route their transactions
through the credit card association. As a result, co-branded transactions are included in
the general-purpose volumes rather than the private-label category.

•

Sub-switching
Sub-switching, also known as private interchange or intra processing, occurs when a
processor is both the issuer and acquirer processor for a transaction. The processor is
able to settle both “sides” of the transaction within their system rather than routing the
transaction through the card organization for settlement. For example, suppose Processor
X is the issuer processor for Bank A and the merchant processor for Merchant B. If a
Bank A cardholder makes a credit card purchase at Merchant B, Processor X could
reconcile that transaction internally since they are linked to both ends of the payment
chain.
Sub-switching does occur, and the largest processors estimate that it accounts for about
9% of their volume. However, per association rules, processors that sub-switch must
report sub-switched volume and pay interchange on it. Transaction volumes reported by
the card associations include these volumes.

This category is the largest source of electronic purchase transaction volume in this study,
accounting for 42% of all transaction volumes in 2000.
Table 35: General-Purpose Credit Card Volumes

Generalpurpose
Credit Cards

•
•

Transactions
(Millions)

Sales Volume
(Millions)

Average
Trans. Size

12,300.2

$1,072,555

$87.20

Annual transactions per card vary significantly, depending upon the type of card.
General-purpose cards are used half as frequently as cards that have been marketed for
more specialized purposes such as T&E.
Consumers also spend significantly more each year on specialized cards.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

5.4.2

Private-Label Credit Cards

5.4.2.1 Background

Unlike the general-purpose credit card industry, where there are just a handful of associations
that needed to be surveyed to collect all relevant data, private-label credit card programs may be
run by individual retailers or gas companies, or by a third party. Generally, larger retailers run
their own programs while smaller retailers outsource this task. In this section the data were
categorized by the owner of the card’s receivables.
•
•
•
•

Retailers
Oil/gas companies
Third-party fleet-card issuers
Third-party receivable owners

5.4.2.2 Organizations

Seventy-two major private label credit card receivable owners were surveyed. Smaller receivable
owners were not surveyed because their individual transactions and sales volumes were too small
to have a significant effect on the aggregate total.
5.4.2.3 Survey Data: Private-Label Credit Cards

Private-label credit card receivable owners were surveyed for totals of U.S. originated approved
and settled transactions for calendar year 2000.
The following information was requested of survey respondents:
•
•

Required data:
o Number and value of transactions processed
Optional data:
o Number and value of transactions by venue (in-store, mail/telephone order,
Internet, other card not present purchases and other card present purchases)
o Number and value of transactions by type of retailer
o Number of active and total private-label credit cards outstanding by type of
retailer
o Number of retailers issuing card by type of retailer

Excluded from the counts provided by survey respondents were:
•
•

Transactions on co-branded cards (accounted for by general-purpose credit card
associations)
Processed transactions for which the respondent does not own the receivables

Private-label credit cards are one-fourth the size of general-purpose credit cards' sales volume in
2000. However, private-label cards still represent a significant source of electronic purchase
transactions in this study, accounting for 9% of all electronic payments in 2000.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

59
Table 36: Private-Label Credit Card Volumes
Transaction
Volume
(Millions)

Sponsored by the Federal Reserve System

60
Exhibit 11: Private-Label Credit Card: Average Transaction Size by Category
$90
$80

$77

$75

$70
$58

$60
$50
$40
$30

$24

$20
$10
$0
R e t a ile rs

5.4.3

O il C a rd s I s s u e rs

3 rd P a rt y F le e t C a rd

3 rd P a rt y

I s s ue rs

R e ce iv a b le O w ne rs

Debit Cards

5.4.3.1 Background

Over the last several years, debit cards have become a more popular method of payment among
consumers. Debit cards can also be used to withdraw cash from ATMs, though those
transactions are not included in this study. Debit purchase transactions can be either online
(PIN-based) or offline (signature-based). While many consumers are unaware of the differences,
there are many aspects of how these transactions are processed that are important to the merchant
and the card issuer.
Table 37: Online vs. Offline Debit Cards
Online Debit Cards

Offline Debit Cards

Can be used at an ATM

Can also be used for purchases at
checkouts with PIN pads (e.g., a grocery
store or gas station) by entering PIN

Can also be used for purchases at checkouts
anywhere Visa/MasterCard is accepted by
signing receipt

Enables cash-back at point-of-sale

Does not allow cash-back

Routed through regional EFT network

Routed through Visa/MasterCard networks

Real-time settlement

ACH settlement in 1-3 days

Typically fixed fee per transaction paid by
merchant to issuer

Fee based on sales amount paid by
merchant to issuer

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

61
5.4.3.2 Online Debit
5.4.3.2.1 Organizations

The data for the online debit payment statistics were gathered from regional and national EFT
networks that provide online debit at the point-of-sale (online POS). All 40 EFT networks were
surveyed to ensure the most accurate count, although 13 of these networks do not offer an online
POS service under their brand. Some networks support the debit POS service provided by
another network; in that case, the transactions was counted by the network that owns the brand.
5.4.3.2.2 Survey Data: Online Debit

EFT networks were surveyed for totals of U.S. originated authorized and settled online POS
transactions for calendar year 2000.
The following information was requested of survey respondents:
•

•

Required data:
o Number and value of purchase transactions at the point of sale that carry the
network brand
o Number and value of purchase transactions at ATMs such as postage stamps
Optional data:
o Number and value of transactions by type of retailer
o Number and value of cash back transactions
o Number of active and total debit cards outstanding by type of card (online,
offline, combined, EBT)
o Number of financial institutions issuing cards by type of financial institution
o Number of merchants, merchant locations, and POS terminals accepting cards by
type of retailer

Excluded from the counts provided by survey respondents were:
•
•
•

ATM transactions (other than as noted above)
Reciprocal or gateway transactions, in order to avoid double counting on the issuer and
acquirer sides of transactions
On-us transactions, where a financial institution is both the issuer and acquirer of a
transaction

As noted above, the survey specified that the online POS category should include only
transactions that carried the network brand, and that reciprocal and gateway transactions should
be excluded. This is because online debit transactions may go through more than one of the
regional or national EFT networks. If a cardholder from one network uses his card at a merchant
location that is affiliated with a different network, then both networks may count that transaction.
Since all transactions carry one and only one network brand, all transactions are counted once
and only once under this methodology.
Due to the number of recent mergers among EFT networks, special care was taken with
information from networks that had acquired or had been acquired during 2000. In these cases,

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

62
full year data were gathered from each organization that was then validated individually with the
sources to ensure there was no double counting of volumes.
Table 38: Online Debit Card Volumes

Online Debit
Cards

Transactions
(Millions)

Sales Volume
(Millions)

Average Trans.
Size

3,010.4

$138,151

$45.89

5.4.3.3 Offline Debit
5.4.3.3.1 Organizations

Currently, Visa and MasterCard have the only two networks for offline debit transactions,
branded Visa Check and MasterMoney respectively. Visa’s offline debit statistics also include
the new hybrid online/offline debit card offered by Visa, the Visa Check Card II. Since both
Visa and MasterCard provide aggregate data, there was no need to survey card issuers, acquirers,
or processors to gather accurate counts. Transactions are routed through one and only one of
these organizations, so there is no risk of double counting.
5.4.3.3.2 Survey Data: Offline Debit

Since these data were collected only from Visa and MasterCard, they were collected on the same
survey form as the credit card data. The form asked for the number and value of approved and
settled offline debit card transactions during the year 2000.
Visa and MasterCard’s offline debit cards can be used anywhere their credit cards can be used;
therefore, the number of merchants and card terminals that accept their offline debit cards is the
same as those for their credit cards.
These numbers are for purchase transactions only, and therefore exclude cash advances.
Table 39: Offline Debit Card Volumes

Offline Debit
Cards

•
•

Transactions
(Millions)

Sales Volume
(Millions)

Average
Trans. Size

5,268.6

$209,980

$39.85

The number of offline debit cards is 135.8 million.
The average cardholder conducts 39 transactions per year totaling $1,546.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

63
5.4.4

Automated Clearinghouse

5.4.4.1 Background38

The ACH Network transfers payments and related data through computer and high-speed
communications technology. ACH Network services can be divided into five broad categories:
•

Direct Deposit Services. Direct Deposit is the automatic deposit of all or part of
employees’ pay, retirees’ pension and annuities, and other business deposits to
consumers’ checking and/or savings accounts.

•

Direct Payment & Home Banking Services. With Direct Payment, consumers
preauthorize electronic debits to their depository accounts for types of recurring bill
payments such as insurance premiums, utility bills, all types of loan payments,
mortgages, club memberships, subscriptions and charitable contributions. Home
banking/bill payment services allow consumers to initiate their bill payments by
telephone, computer, or other mechanisms.

•

Electronic Commerce. Electronic commerce can incorporate all aspects of the ordering,
inventory, and payments processes of businesses. Companies may use electronic data
interchange (EDI) to place orders; to update, control, and reorder inventories; to transmit
billing statements; and to collect or make payments.

•

Electronic Benefits Transfer (EBT). EBT enables governments to replace multiple
paper systems with a single, streamlined electronic delivery system that delivers benefits
for a wide range of Federal and state programs. Most frequently, EBT is used to provide
food stamp and cash assistance benefits. EBT allows recipients to access their benefits
with the use of a card through automated teller machines (ATMs) and retail point-of-sale
(POS) terminals.

•

Electronic Checks. An electronic check is the conversion of a paper check to an
electronic transaction as early as possible in the check collection process. Another
electronic check concept is Electronic Check Presentment (ECP). Financial institutions
use ECP to send an electronic file of MICR-line information to the paying bank in
advance of the paper check. The paper check continues through the transaction
processing and, in due course, is physically received at the paying bank.

5.4.4.2 Organizations

Four ACH network operators handle all ACH transactions (except on-us transactions, described
below). The data for ACH statistics were gathered from these four operators. Additionally, the
National Automated Clearing House Association (NACHA) collects annual statistics from these
networks as well; NACHA’s data for 2000 were used to validate and verify respondents’
information.

Condensed from the 2001 NACHA Buyer’s Guide

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

64
5.4.4.3 Survey Data: Automated Clearing House

The ACH operators and NACHA were surveyed for totals of U.S. originated transactions for
calendar year 2000, excluding ODFI/RDFI settlement transactions.
The following information was requested of survey respondents:
•
•

Required data:
o Number and value of ACH debit and ACH credit transactions and returns for
originated inter- and intra-operator transactions by SEC code.
Optional data:
o Number and value of ACH debit and ACH credit transactions and returns for
originated intra-operator transactions by SEC code.
o For operators, number of ODFIs that they receive data from, and number of
RDFIs that they send items to
o For NACHA, estimated number of on-us transactions processed by originators

Table 40: Automated Clearing House Volumes

Automated
Clearing House

Transactions
(Millions)

Dollar Value
(Millions)

Average Trans.
Size

5,622.0

$5,674,851

$1,009.40

All ACH transactions go through one or more of the four ACH operators. Similar to online debit
transactions, a risk of double counting exists when a transaction is routed through more than one
operator. To avoid any potential double counting, data were collected on originated transactions
only. That is, operators were asked to include only transactions that originate within their own
network (whether it is transmitted to another operator or is transmitted to the RDFI).
5.4.4.4 ACH Transaction Classifications

There are several ways an ACH transaction can be classified. These are described below:
5.4.4.4.1 Network vs. On-Us Transactions

In addition to ACH network volume that goes through one of the ACH operators, ACH
transactions may occur “on-us” at the financial institution level. An example of this would be a
case where a single DFI is both the ODFI and the RDFI for a given transaction. The DFI would
not need to route the transaction through one of the ACH operators.
Counting the entire population of on-us transactions would require surveying all financial
institutions that use ACH. NACHA surveys the 50 largest financial institutions to form an
estimate of total on-us transaction volumes. Those estimates were added to the network data
collected from the operators.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

5.4.4.4.2 Originated vs. Received Transactions

An ACH operator can originate or receive a transaction. All ACH transactions are originated
through one operator, and they may or may not be passed on to another “receiving” operator. If
the transaction is passed on, and both operators count it, then it would be double-counted. To
avoid this, operators were asked to provide data on originated transactions only.
5.4.4.4.3 Inter- vs. Intra-Operator Transactions

ACH operators also classify their transactions as inter- or intra-operator transactions. Interoperator transactions are those that are transmitted to another ACH operator. Intra-operator
transactions are those that are not transmitted to another ACH operator. This would be the case
if the operator were the ACH operator for both the ODFI and the RDFI. This survey wanted to
capture ACH transactions regardless of whether they went through one or more operators (noting
the risk of double-counting mentioned above), so operators were asked to provide a combined
total of inter- and intra-operator volumes.
5.4.4.4.4 Debits vs. Credits

All ACH transactions are classified as an ACH debit or an ACH credit, depending on whether
the originator is crediting an account or debiting an account. Either of these is considered a
transaction, so they are combined to an aggregate total for purposes of this study.
5.4.4.4.5 Returns

Like a credit card or debit card transaction, ACH transactions can be returned. However, the
reporting of returned transactions is more complex within the ACH system. ACH operators
report returns differently.
EPN and the Fed report the number of returns by SEC code, and these can be subtracted from
each SEC code’s transaction volumes. In cases where the original SEC code was not available,
the return is reported under the “RET” SEC code. These operators do not report dollar value of
returns. To estimate these volumes, the average transaction size for each SEC code was
calculated by dividing dollar value by transactions, and multiplied by the number of return
transactions.
ACHA and Visa report total returns only; they do not provide this data by SEC code. However,
they do provide both the number and dollar value of returns. To estimate the returns by SEC
code, the share of transactions by SEC code was calculated by dividing each SEC code’s
transactions by the total transactions, and applying this percentage to the total returns. The same
was done on the dollar value side. These estimated returns were then backed out of each SEC
code.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

5.4.4.4.6 SEC Codes

All ACH transactions are routed under one of several Standard Entry Class (SEC) Codes defined
by the ACH operating rules. There were 21 such codes effective during the year 2000, though
no data were reported under four of the codes. About half of the active codes relate to domestic
payment transactions, and are therefore included in this study. The other half concern crossborder transactions, ATM withdrawals or simple notifications, and are therefore excluded from
our totals. A list of which SEC codes are included and excluded from the study results is shown
in the following table.
Table 41: SEC Codes Included in ACH Aggregates
Code
CIE
CTX
POS
PPD
POP
RCK
SHR
TRC
TRX
XCK

Description
Consumer Initiated Entry
Corporate Trade Exchange
Point of Sale Entry
Prearranged Payment and Deposit Entry
Point-of-Purchase Entry
Re-presented Check Entry
Shared Network Transaction
Truncated Entry *
Truncated Entries Exchange *
Destroyed Check Entry

* Inactive code

Table 42: SEC Codes Excluded from ACH Aggregates
Code
ACK
ADV
ATX
CBR
CCD
COR
DNE
ENR
MTE
PBR
RET

Description
ACH Payment Acknowledgement
Automated Accounting Advice *
Financial EDI Acknowledgement *
Corporate Cross-Border Payment
Cash Concentration or Disbursement
Automated Notification of Change
Death Notification Entry
Automated Enrollment Entry
Machine Transfer Entry
Consumer Cross-Border Payment
Automated Return Entry

* Inactive code

The network totals calculated from the survey responses were about one percent different from
the data published by NACHA. This difference may be accounted for by the fact that some of
the data collected from the operators had been revised since they were given to NACHA.
Table 43: ACH Transactions By SEC Code
SEC Code
CIE
CTX

Transactions
(Millions)
36.3
33.6

RETAIL PAYMENTS RESEARCH PROJECT

Dollar Value
(Millions)
$24,498
915,381

Average Trans.
Size
$675
27,260

Sponsored by the Federal Reserve System

67
POP
POS
PPD
RCK39
SHR
XCK
Total

14.6
29.8
5,467.5
2.9
37.2
0.1
5,622.0

1,088
1,607
4,729,293
483
2,472
28
5,674,851

75
54
865
165
67
245
1,009

As shown in the table, the PPD code is the largest by far, accounting for 97% of the transactions
and 83% of the dollar value. The PPD code (for Prearranged Payment and Deposit Entry)
includes such items as direct deposit of payroll, pre-authorized bill payment and check
conversion.
Also notable is that different SEC codes can have dramatically different average transaction
sizes. This is due to the different nature of the transactions routed under each code. For
example, CTX transactions are corporate exchanges from one organization to another, and
therefore are very high. POS transactions are individual purchases at the point of sale and are
therefore relatively small.
5.4.5

Electronic Benefit Transfer

5.4.5.1 Background40

Electronic Benefit Transfer (EBT) is an electronic system that allows a recipient to authorize
transfer of his/her government benefits from a Federal account to a retailer account to pay for
products received. EBT is currently being used in many states to issue food stamps and other
benefits. Nearly 80 percent of food stamp benefits are currently being issued by EBT.
EBT accounts are established in the participant’s name, and food stamp and other benefits are
deposited electronically in the account each month. A plastic card, similar to a bank card, is
issued and a personal identification number (PIN) is assigned or chosen by the recipient to give
access to the account.
States are steadily bringing new EBT projects on line. As of September 2000, forty states, the
District of Columbia, and Puerto Rico were using EBT in some form to issue food stamp
benefits: The remaining states are in various stages of planning for EBT. The Welfare Reform
Act of 1996 mandates that all states must switch to EBT issuance for the Food Stamp Program
by October 2002. Although this act only applies to food stamp benefit distribution, more and
more states are adding cash benefits to their EBT system. Cash benefits can, depending on the
state, be withdrawn at an ATM or spent at a retailer with a POS terminal.

The RCK volume and value are considered electronic payments for the purposes of this study. It should be noted,
however, that for each RCK transaction the original intent of the payor and payee was to settle using the check
system. The inability to settle via the check system caused the check to be collected via ACH.
40
Condensed from the Food & Nutrition Service’s Website

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

68
5.4.5.2 Organizations

All states participating in EBT have a single primary contractor that administers their EBT
payments program. That contractor may subcontract processing or any other aspect of the
program to another company. Because of these complex relationships, and to ensure that no
transactions were double counted, only the primary contractors were surveyed. There are
currently only four primary EBT contractors.
In addition to these four contractors, we surveyed the Food & Nutrition Service (FNS). FNS
collects data on food stamp benefits administered through EBT programs, but does not have data
on cash benefits that may also be administered through EBT programs.
5.4.5.3 Survey Data: Electronic Benefit Transfer

The EBT contractors were surveyed for total approved and settled purchase transactions for
calendar year 2000.
The following information was requested of survey respondents:
•

Required Data
o Number and value of purchase transactions processed

•

Optional Data
o Number and value of transactions by type of retailer
o Number of active and total EBT cards outstanding
o Number of merchants, merchant locations, and POS terminals accepting cards by
type of retailer

Excluded from the counts provided by survey respondents were:
•

ATM transactions

EBT transactions can be broken down between food stamp benefits, which all states participating
in EBT provide, and cash benefits, which some states provide. Cash benefits are primarily
through Temporary Assistance for Needy Families (TANF), which replaced Aid to Families with
Dependent Children (AFDC) in 1996, but may also include other state programs. Similar to a
bank account with a debit card, these cash benefits can be accessed in two ways: as cash from an
ATM, or as a purchase made at a retailer by swiping the card and entering a PIN. For purposes
of this study, we are counting all food stamp benefits, since these are accessed by making a
purchase, as well as the portion of cash benefits that are accessed as a purchase rather than a cash
withdrawal.
Statistics for this category were gathered on the purchasing side rather than the funding side.
That is, when an EBT cardholder makes a purchase (whether using food stamp benefits or cash
benefits), that transaction is counted here. Funding transactions, including transfers from the
state or federal government to financial institutions that hold funds, and transfers from those fund

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

69
pools to merchants who accept EBT cards, are not counted here. Funding transactions go
through the ACH system and are included in those numbers.
This results in the same funds being counted more than once. In fact in this case, the same funds
are counted three times: first as they are transferred from the government to the financial
institutions, again when they are transferred from the financial institutions to the merchants, and
a third time on the purchasing side when consumers make purchases on their EBT cards.
However, each of these funds transfers is a separate event, and represents an opportunity for an
electronic funds transfer to replace a paper funds transfer. Therefore, each transaction is
counted. In the past, these would have been three separate paper transactions.
Direct Federal payments to individuals are made through the Financial Management Service
(FMS) of the Department of the Treasury. These payments include Social Security, SSI,
Veterans, and other government programs. These transactions are authorized and executed
through the Fed ACH system and are counted, therefore, with the ACH numbers.
Table 44: Electronic Benefit Transfer Volumes

Electronic Benefit
Transfer

Transactions
(Millions)

Sales Volume
(Millions)

Average Trans.
Size

537.7

$13,744

$25.56

All but two states use a magnetic-stripe card for their EBT program. Ohio and Wyoming
implemented a chip card program. Additionally, Nevada is piloting a program to offer WIC
benefits on a chip card.
5.4.6

Emerging Payment Technology Companies

5.4.6.1 Background

In recent years several new payment technologies have been developed. However, these
products are generally just new front-end payment methods to the consumer, using traditional
funding and settlement systems behind the scenes. Therefore, these types of transactions are
counted within the basic funding and settlement systems that they use and are not separately
added into the aggregates.
We surveyed a number of these companies to get a sense of the volume that they are
contributing. Very few of these companies responded to our survey. In many cases, it was
simply because the company was so new that it did not yet have any significant volume. In other
cases it was because they are in a new and competitive market and do not want to reveal their
transaction data. Still others did not respond simply because they are a small operation and did
not have the time or manpower to respond to a survey.
We will continue attempting to quantify these emerging payment technologies in future studies.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

70
Emerging payment technology companies can be categorized as follows:
•

Bill Payment Companies. Electronic bill payment and presentment (EBPP) refers to
online services that enable customers to receive, review and execute payment of their
bills over the Internet.

•

Person-to-Person Payments. A person-to-person (P2P) payment involves an
electronically initiated transfer of value from one individual to another. Using PCs,
handheld computers and mobile phones, individuals can use this anytime, anywhere
service to send money to family members, settle debts with friends and pay for items
purchased through online auctions.

•

Stored Value. Prepaid cards are promoted for a number of uses. Though perhaps they
are currently best known for their gift card application, as a replacement for a gift
certificate, they are also being used for payroll, incentives, insurance, refunds and other
purposes.

•

Internet Currencies. Internet currencies are, as the name implies, currencies intended to
be spent on the Web. Web merchants must be set up to accept an Internet currency, and
they are generally not widely accepted, though some are much more popular than others.
Some can also have their value transferred to a card and spent at a physical location.

•

Other Emerging Payment Technologies. There are several other types of emerging
payment technologies as well.
o Several companies are working on ways to allow consumers to use their online
debit cards for Internet purchases. These technologies route transactions through
the EFT networks.
o Transponders allow consumers to waive a small tag in front of a reader to pay for
goods. Purchases paid for with a transponder are billed to the consumer’s credit
card or to a prepaid account.
o A major processor is working with a supermarket industry trade group to pilot an
ACH debit card, which works similarly to an online debit card, but routes
transactions through the ACH system rather than an EFT network.
o Several companies have developed ways to convert checks to electronic
transactions at the point-of-sale. These transactions would then go through the
ACH system.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

Appendices

APPENDICES

6.1

Appendix A – Statistical Methodology for The DFI Check Study

The DFI Check Study received check volume and value data from 1,256 financial institutions
during March and April 2001.
6.1.1

Sample Design

The sample for The DFI Check Study was based on a stratified single-stage design with
systematic sampling of DFIs (i.e., holding company if applicable) using a random start. The
primary strata were defined on the basis of type of institution, that is, commercial banks (CMB),
credit unions (CUS) and thrifts (THR).41 The next level of stratification was carried out on the
basis of size where the measure of size was the public checkable deposits (PCD) value at the
highest institutional level (e.g., holding company). Public Checkable Deposits (PCD) are all
checkable deposits held by a DFI that are not the deposits of other DFIs or the federal
government. Checkable deposits were believed to be a better indicator of check volume than
total assets or total deposits.
The sampling unit was the DFI at its highest institutional level (e.g., holding company) and the
data were collected for all the institutions owned by the sampled institution.
6.1.1.1 Sampling Frame of the Financial Institutions

The sampling frame was constructed from files supplied by the Federal Reserve Board of
Governors. The frame represented the population of insured depository financial institutions in
the United States, which includes U.S. branches of foreign owned institutions. Only institutions
with checkable deposits above $100,000 were included in the frame. It is possible that a bank
holding company could have no checkable deposits, in which case it would be eliminated from
the frame.
More specifically, the frame consisted of:

6,846 commercial banks and bank holding companies, plus 6 "anomalous banks"
6,551 credit unions,
1,293 thrifts.

The six anomalous banks were identified and surveyed as a certainty stratum, because their paid
check volume was known to be poorly correlated to PCD. Relatively speaking, these were small
41

Thrifts include savings banks and savings and loan associations.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

72
banks (low PCD value) that process a high volume of low-dollar value rebate checks. These
institutions were surveyed as a certainty stratum to avoid the risk of selecting them as part of a
random sample. They are not representative of most other institutions their size. Their data
would skew the results of a national estimate if it were extrapolated to estimate volumes
processed by non-sampled banks.
The frame also contained numerous variables, some of which were used in stratification, and
some of which were used as identifiers. Public checkable deposits (PCD) was the stratifying
variable.42 Variables used for identification, but not stratification were bank name, city, state,
ID_RSSD, ID_TOP and ABA_RTN.43 The identifier variables were used to map institutions to
appropriate data sources for mailing lists, for example, Thomson’s Financial Databank, an
industry directory of the banking industry.
6.1.1.2 Stratification of Sampling Frame

The primary strata were based on the type of institution: CMB, CUS, and THR. This means that
all of the CMB constituted a stratum, all CUS another stratum, etc. Within the primary strata (the
CMB, CUS, and THR) there was further stratification by size. For instance, the largest
institutions were in one stratum, the next largest in another, all the way to the very smallest in the
last stratum.
Because the largest institutions account for the majority of total paid check value and volume,
the 100 largest institutions were included with certainty. All institutions in the certainty stratum
were included in the sample. Sampling was conducted to select institutions from the other strata.
The total sample of 2,339 institutions was allocated across 14 design strata defined by type of
institution and size (plus one stratum of anomalous banks).
6.1.1.3 Sample Size and Sample Allocation

The survey estimates were based on assumptions for each of the primary strata (CMB, CUS and
THR), and the aggregates of these strata:

Expected response rate of 65 percent.
Sample allocated to achieve less than a 10.0 percent confidence value (CV) for
commercial banks, credit unions and thrifts for the size measure under the assumption of
a 65 percent response rate within each stratum.

Since survey estimates were produced both at the primary stratum level (type of institution) and
at the aggregate level the criteria were also applied so that the sample allocation across primary
strata should result in certain desired CVs for the primary strata. As discussed earlier, the
primary strata were the three types of institution and the secondary strata were defined on the
42

The CUS list included a share draft field, S_DRAFT, as the stratifying variable. For simplicity, however, we refer
throughout the report to PCD as the stratifying variable for all DFIs.
43
ID_RSSD is an identification number the Federal Reserve assigns to every entity listed in its database (including
DFIs, DFI branches, bank holding companies, etc.). It stands for IDentification_Research, Statistics, Supervision,
and Discount. The ID_TOP for a bank is the ID_RSSD of the highest level holding company that owns that bank. It
stands for IDentification of the TOP Holder. ABA_RTN is the Routing Transit Number for the institution.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

73
basis of a measure of size (PCD). The desired CV of 10 percent or less was obtained under the
assumption of a uniform response rate of 65 percent within the secondary strata including the
TAKE-ALL strata.
The Lavallee and Hidiroglou (1988) procedure, which defines optimum TAKE-ALL (certainty)
and TAKE-SOME (sampling) strata boundaries, was used to determine the minimum sample
sizes. The procedure minimizes the total sample size while achieving the required relative
standard error. Also, by using the Lavallee and Hidiroglou (1988) procedure, we not only
determine the required sample sizes, but we also obtain the size stratification within each
primary stratum. The number of size strata was 5 for each of the commercial banks and the credit
unions, and 3 for the thrifts. Thus, the total number of design strata was 13 out of which 10 were
TAKE-SOME (or sampling) strata. The TAKE-ALL stratum for the commercial banks was
further divided into 2 TAKE-ALL strata for operational reasons. Moreover, this would also
result in more homogeneous nonresponse adjustment classes as discussed later.
By applying the above procedures and taking into consideration the expected response rate of 65
percent, we arrived at a total sample of 2,339 institutions. The table below gives the number of
institutions in the sample frame and the number sampled by sampling stratum. There are 14
sampling (or design) strata. Out of the 14 design strata, the sampling was conducted in 10 strata
only and the remaining 4 were defined to be TAKE-ALL (or certainty) strata.
Table 45: Number of Institutions Sampled by Size Stratum (Original Design)
Primary
Stratum
Commercial
Banks

Credit
Unions

Thrifts

Size
Stratum
1
2
3
4
5
6
All
1
2
3
4
5
All
1
2
3
All

TOTAL

Number of
Institutions
204
329
845
1,408
2,036
2,008
6,830
104
344
723
1,742
3,199
6,112
40
347
850
1,237
14,179

Number
Sampled
204
329
336
300
242
156
1,567
104
181
134
112
69
600
40
114
44
198
2,365

6.1.1.4 Sample Selection

The sampling was performed at the highest institutional level. Where applicable, therefore, the
sampling unit was the holding company. The sample was selected using a systematic sampling

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

74
procedure within each sampling stratum. The institutions in the sample frame were first sorted by
size within each of the design strata for systematic sampling.
6.1.1.5 Selection Probabilities

The sample design for the survey of financial institutions is a single stage stratified design with
systematic sampling of institutions from within strata. The selection probabilities for all units
within a stratum are equal to the sampling rate for the stratum. The selection probabilities are
equal to 1 for all institutions in the TAKE-ALL strata.
Let h denote the stratum and the index i denote an institution (e.g., holding company) within the
stratum. In order to keep the notation simple, we will use the index h to denote the final design
strata obtained after the size stratification. If we select nh (the sample size for stratum h) out of
the N h (the population size for stratum h) institutions in stratum h, then the selection probability
for stratum h, say π h is given by

π h = nh N .
h
The selection probabilities π h define the base weights as discussed in Section 6.1.2.1.
6.1.2

Sample Weighting

6.1.2.1 Base Weights

Survey responses from the respondents are inflated to obtain estimates for the entire population
using sampling weights. These weights are designed to (1) compensate for unequal selection
probabilities; (2) attempt to compensate for unit nonresponse, that is, nonresponding institutions;
and (3) attain greater precision for the survey estimates through poststratification (or ratio
estimation) if auxiliary data for poststratification are available. As discussed in Section 6.1.1.5.,
the selection probabilities of the institutions will vary by stratum due to different sampling rates
for the strata. The base weight for a sampled institution is defined to be the reciprocal of
sampling rate for the corresponding stratum, which is also the selection probability of the
institution. Such base weights produce unbiased estimates of totals and percentages in the
absence of any nonresponse in the survey.
If π h is the selection probability for the sampled institutions from stratum h then the base weight
(or design weight) assigned to the corresponding institutions is defined as
Wh = 1 π h .
Properly weighted estimates using the base weights (as given above) would be approximately
unbiased if every sampled unit agreed to participate in the survey. However, nonresponse is
always present in any survey operation. The DFI Check Study was no exception; a number of
sampled institutions declined participation in the study. Thus, weight adjustments were
necessary to minimize potential nonresponse bias.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

75
Ratio adjustment of sample weights at the individual stratum level was also performed using the
auxiliary variable (PCD) to reduce the variability of sample estimates. The improvement in the
reliability of the estimates due to ratio adjustment depends on the correlation between the study
variable and the auxiliary variable used for ratio adjustment.
6.1.2.2 Nonresponse Adjustment

The base weights are adjusted to account for the nonresponding institutions. As described in
Section 6.1.2.1. the base weights would produce approximately unbiased estimates only if there
were no nonresponding institutions. Due to the presence of nonresponse, a weight adjustment
was required to account for nonresponding institutions. This weight adjustment was obtained by
increasing the weights of the sampled institutions for which data were collected. The
nonresponse weight adjustment was applied within each stratum. Suppose that nh institutions
were sampled from stratum h and data were collected for rh institutions only. Then the
adjustment is calculated as the ratio of the number of sampled institutions (both respondents and
nonrespondents) in the stratum to the number of institutions that actually responded to the
survey. Let Ah(nr ) denote the weight adjustment due to the nonresponding institutions within
stratum h, then
Ah( nr ) =

In order to compensate for the nonresponse, the base weight for stratum h, that is, Wh , was
multiplied by the nonresponse adjustment factor Ah(nr ) given above. The stratum level design
weight adjusted for nonresponse then becomes

Wh* = Ah( nr )Wh .
The above weight can be assigned to all institutions belonging to stratum h for which survey data
were obtained. Although, above weights can be used to produce unbiased survey estimates, we
also used the auxiliary information on the size of institutions from the sampling frame for the
purpose of ratio estimation. The benefit of ratio estimation is that it helps reduce the variability
of survey estimates.
6.1.2.3 Ratio Estimation

Ratio estimation is a popular estimation procedure in which the weights of the respondents are
further adjusted so that the survey estimates of an auxiliary variable are equal to the known
population totals of the auxiliary variable for each stratum or some aggregate of strata. We
applied the ratio adjustment at the stratum level because the sample sizes at the stratum level
were adequate.
Let X h denote the sum of the PCD values for all the institutions in the stratum denoted by h

(h = 1,2,− − −, L) as obtained from the sampling frame, and let X̂ h be the corresponding survey

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

76
estimate obtained by using the nonresponse adjusted base weight. Then the ratio

is used
Xˆ h

as an adjustment to define the ratio estimation weight Wh(ratio ) as

X 
Wh( ratio ) =  h Wh* .
ˆ
 Xh 
The superscript (ratio) denotes the ratio estimation weight and the weight is applied to all
institutions in stratum h. It can easily be shown that the ratio estimation weight Wh(ratio ) for
stratum h can also be expressed as
N
Wh( ratio ) =  h
 rh

 Xh

,
 xh

where x h and X h are respectively the sample mean and the population mean of the auxiliary
variable PCD for the stratum denoted by h, and rh is the number of responding institutions from
stratum h. The ratio estimation weights Wh(ratio ) are the final survey weights and these are used to
tabulate the survey responses.
6.1.3

Estimation

The survey covers three types of financial institutions, commercial banks (CMB), credit unions
(CUS) and the thrifts (THR). Both aggregate level and type of institution level estimates were
produced from The DFI Check Study. In this section we discuss the estimation procedures for
these two levels of survey estimates. For the sake of simplicity we will denote by Wh the ratioadjusted survey weight for stratum h instead of Wh(ratio ) as defined in section 6.1.2.3. Also, we
will use nh instead of rh to denote the number of respondent institutions from design stratum h.
6.1.3.1 Estimates of Totals

The form of the survey estimate for the characteristic y at the stratum level is given by
nh

Yˆh = ∑ Wh y hi
i =1

where h denotes the design stratum, and i is the sampled (and respondent) institution from
stratum h. The variable Wh denotes the sampling weight for stratum h and yhi is the observed
value of the variable (or characteristic) y for the responding institution i from the stratum h.
The estimate of the stratum level total of the characteristic y can also be expressed in an alternate
form as

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

y 
Yˆh = N h  h  X h
 xh 

where y h is the sample mean of the study variable y (or characteristic of interest).
Then the estimate of the aggregate of characteristic y is simply the sum of the estimates of the
individual design strata, that is,
L

Yˆ = ∑ Yˆh
h =1
L

=∑

∑W

h =1 i =1

y hi

where L is the total number of design strata.
The estimate of the total of characteristic y by type of institution is simply the sum of the
estimates of the individual design strata belonging to that type, that is,
Yˆ( CMB ) =
=

∑ Yˆ

h∈CMB

h
nh

∑ ∑W

h∈CMB i =1

y hi

where summation is over those strata that belong to the commercial banks. Similarly, the
estimates for the credit unions and the thrifts can also be obtained by summing the stratum level
estimates over the strata for the corresponding types.
6.1.3.2 Reliability of the Estimates

Because estimates are based on sample data, they differ from figures that would have been
obtained from complete enumeration of the universe using the same instruments. Results are
subject to both sampling and nonsampling errors. Nonsampling errors include biases due to
inaccurate reporting, processing and measurement, as well as error due to nonresponse and
incomplete reporting. These types of errors cannot be measured readily. However, to the extent
possible each error has been minimized through the procedures used for data collection, editing,
quality control and nonresponse adjustment.
The sampling error (or standard error) of an estimate is defined as the square root of the variance
of the estimate. The standard error of an estimate is inversely proportional to the square root of
the number of observations in the sample. Thus, as the sample size increases, the standard error
decreases.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

78
6.1.3.3 Variance Estimation of Estimates of Totals

The estimates of variances of the estimated totals from the survey have been computed by
employing Taylor Series approximations (or Taylor linearization). The Taylor linearization
technique has been used because the estimates of totals have been obtained using ratio estimation
weights. Wolter (1985) is a useful reference on the Taylor Series method of variance estimation.
As described earlier, the weights were computed as separate ratio estimation weights at the
design stratum level using auxiliary variable PCD for ratio estimation. The estimate of the
aggregate of characteristic y is simply the sum of the estimates of the individual design strata,
that is,
L

Yˆ = ∑ Yˆh
h =1
L

=∑

∑W

h =1 i =1

y hi

where L is the total number of design strata.
The variances are obtained at the design stratum level and the variances of the aggregates are
simply the sum of the stratum level variances. In order to obtain the variance of the stratum level
total Ŷh we define synthetic variable d hi as
d hi = y hi − Rh x hi
where Rh is the ratio of the stratum level average value of the study variable y to the stratum
level average value of the auxiliary variable. The ratio Rh is unknown, and was substituted by its
y
estimate defined as h , where y h is the sample mean of the study variable y and x h is the
xh
sample mean of the auxiliary variable “PCD.” The variance of the estimated stratum level total
Yˆh is then given by
X
var(Yˆh ) = (1 − f h ) ×  h
ˆ
 Xh
where f h =

2
nh
 N h2
1

(d hi − d h ) .
 n (n − 1) ∑
i =1
 h h

is the sampling fraction in stratum h and d h is defined as

dh =

RETAIL PAYMENTS RESEARCH PROJECT

1
nh

∑d
i =1

Sponsored by the Federal Reserve System

It should be noted that nh is the number of responding institutions from stratum h and not the
number initially sampled.
The variance estimate of the estimated total is simply the sum of the individual strata level
variances. Thus, the variance of the estimated total Ŷ is given by
X
var(Yˆ ) = ∑ (1 − f h ) ×  h
ˆ
h =1
 Xh
L

2
nh
 N h2
1

(d hi − d h ) .
 n (n − 1) ∑
i =1
 h h

6.1.3.4 Construction of Confidence Intervals

Each of the survey estimates will have an associated standard error, which is defined as the
square root of the variance of the estimate. Let Yˆ be the estimated total of the study variable y
and var(Yˆ ) be the corresponding variance estimate. Then the standard error of the estimated
total Yˆ is given by
sd (Yˆ ) = var(Yˆ ).

The 95 percent confidence interval is the interval such that there is a 95 percent chance that the
unknown total Y would be within the interval. The 95 percent confidence interval is defined as

Yˆ ± 1.96 × sd (Yˆ ).

The lower limit of the interval is
Yˆ − 1.96 × sd (Yˆ ).

And the upper limit of the interval is
Yˆ + 1.96 × sd (Yˆ ).

The width
1.96 × sd (Yˆ )
is known as the half-width of the 95 percent confidence interval. The smaller the half-width of
the confidence interval, the more precise is the survey estimate.
The estimates and the corresponding 95 percent confidence intervals were obtained for two study
variables, Volume and Value, using the months of March and April 2001. The estimates and the
confidence intervals were obtained at both the primary stratum level and the aggregate level. The
results of the estimation are given in Table 9 and Table 10.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

80
6.1.3.5 Restratifying with the New PCD Data

As discussed earlier, the initial sample design was based on a stratified single-stage design with
systematic sampling of institutions using a random start. The primary strata were defined on the
basis of type of institution, that is, commercial banks (CMB), credit unions (CUS) and thrifts
(THR). At the conclusion of data collection, the final data set was restratified prior to analysis to
more accurately reflect both stratification variables. For the primary stratification variable, type
of institution, adjustments were made to reclassify several holding companies that owned both
commercial banks and thrifts (dual-type holding companies). These cases had come to the
attention of the study team either from data the institutions had reported during the study, or
from a review of auxiliary data sets maintained by the Federal Reserve. A decision rule was
employed that reclassified each dual-type holding company into the institution type stratum
containing over 50 percent of their PCD values. So a dual-type holding company owning a
commercial bank (or banks) with PCDs at 75 percent of the total PCDs for all owned institutions
was classified as a commercial bank, regardless of where that holding company was classified in
the original design. A total of 131 institutions were moved from their initial type of institution
stratum - 91 originally classified as commercial banks were reclassified as thrifts, and 40
originally classified as thrifts were reclassified as commercial banks.
The next level of stratification for the design was done on the basis of size where the measure of
size was the PCD value at the highest institutional level. The initial sample design was based on
PCD from 3rd quarter 2000. At the conclusion of data collection (1st quarter 2001) PCDs were
available, which reflect the status of participating institutions during the same time period as the
check volume and value totals reported on the survey. The 2001 PCDs were used as the size
measure for the final design.
We used the PCD value as stratification variable for The DFI Check Study design. The
information is also used for estimation purposes. We received in August 2001 files from the
Federal Reserve Board with updated PCD values.
As these new PCDs were being implemented, some additional modifications to the size strata
within type of institution were made to improve design efficiency. These final improvements
involved adjusting the certainty strata for each type of institution stratum to obtain certainty
groupings based on whether institutions responded to the survey. For commercial banks, the
initial design included two certainty strata containing 533 institutions. In the adjusted design, the
eighth largest commercial bank holding company did not respond, so a stratum containing the
seven largest commercial bank holding companies was used as the only certainty stratum, with
weights equal to 1. A similar method was used for credit unions (obtained certainty stratum with
2 institutions) and for thrifts (obtained certainty stratum with 3 institutions). For the analysis,
design efficiency was improved by excluding each certainty stratum from the computation of
sample variation.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

81
Table 46: Final Depository Financial Institution Check Study Sample Information
Stratum

PCD Range
($ thousands)

Certainty

56,759,826 15,385,621
9,927,979 244,559
242,032 96,461
96,271 47,302
47,296 23,136
23,123 9,442
9,442 105
100 - 1

SAMPLE
Commercial
Banks (CMB)

1
2
3
4
5
6
Below
Minimum*
TOTAL
Credit Unions
(CUS) **

Certainty
1
2
3
4
5
Below
Minimum*
TOTAL
Certainty

Thrifts (THR)
1
2
3
Below
Minimum*
TOTAL
Anom. Banks

Certainty

1,643,158 798,356
431,679 75,062
74,898 25,880
25,843 8,211
8,201 2,008
2,006 101
100 - 1
5,148,010 2,810,146
2,107,059 189,311
187,228 28,296
28,282 115
93 - 3

n/a

All Institutions

Sample
Frame

Institutions Institutions Response Sampling
Sampled *** Responding
Rate
Weight
7

100%

1.000

166

165

113

68%

1.152

329

330

172

52%

1.918

846

333

175

53%

4.769

1,399

315

153

49%

8.804

2,041

246

116

47%

17.361

2,008

151

49%

25.270

6,846

1,547

810

52%

100%

1.000

104

73%

1.315

343

170

111

65%

3.065

721

140

56%

7.735

1,742

111

49%

32.900

3,192

31%

150.302

447

6,551

599

343

57%

100%

1.000

41%

2.430

346

108

56%

4.805

894

46%

34.302

1,293

187

52%

100%

1.000

14,696

2,339

1,256

54%

* In all 3 sample groups, institutions with a PCD of $100,000 or below were not included for sampling.
** For credit unions, the PCD values are the S_Draft (i.e., share draft) variable from the sample frame file.
***The totals in this column are adjusted for merged institutions.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

82
6.1.3.6 Imputation

Prior to final weighting and preparing estimates from the study data, missing data were imputed
for both aggregate and clearing method data fields. The imputation rules were different for the
two classes of response.
6.1.3.6.1 Imputation Methodology for Aggregate Data

The aggregate total monthly volume and value data cases were identified for imputation based on
giving at least one aggregate response. So if volume or value was given for either month (e.g.,
value for was given for both months or volume for the other month was missing), the missing
values were imputed. No case with all four data elements blank (missing) was imputed. Missing
aggregate data were imputed using regression imputation methodology. The linear regression
models were fitted separately for the two data collection moths (March and April 2001) and the
three types of institutions for each value and volume. Thus, six regression models were fitted for
imputing the missing data for the variable check volume (one regression model for each of the
three institution types for each of the two months). Similarly, six regression models were fitted
for imputing the missing data for the variable check value. The PCD value was used as the
dependent variable in the regression model. Each regression model was fitted twice, first to
detect the outliers and then to re-estimate the same model by removing the outliers, which was
used for imputation. The criteria for the outlier detection was that the normalized residuals from
the regression were greater than 3.0 in absolute value. Although, the outliers were not used to
estimate the model for the purpose of imputation, these outlier values were not imputed either.
The goodness of the model fit was judged from the R 2 value of the model. These values for the
different models are reported in Table A.
Table A: R 2 values for different models

Institution Type

Month

Variable

R 2 Value

Commercial Banks
Commercial Banks
Commercial Banks
Commercial Banks
Credit Unions
Credit Unions
Credit Unions
Credit Unions
Thrifts
Thrifts
Thrifts
Thrifts

March
March
April
April
March
March
April
April
March
March
April
April

Volume
Value
Volume
Value
Volume
Value
Volume
Value
Volume
Value
Volume
Value

91.1 %
98.7 %
87.8 %
98.5 %
92.3 %
95.3 %
80.7 %
96.4 %
88.3 %
78.1 %
88.2 %
77.2 %

As we can note from the above table, the regression model fit is better for Commercial Banks
and Credit Unions as compared with Thrifts. Moreover, the regression model did not have as

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

good fit for Value in April as for the other Thrift variables, i.e., March Value and Volume, and
April Volume. Nevertheless, the regression models were adequate for imputing the missing data.
6.1.3.6.2 Imputation Methodology for Clearing Method Data

For detail data about paid check volume/value allocated by clearing method, the criterion to do
imputation was respondent completion of any of the clearing method data fields or of returned
check data fields. In addition, the following rules were used to guide imputation decisions on
each separate data field. The same rule was applied to volume and value for each variable (e.g.,
Federal Reserve Presentments).
1. If a respondent returned the short form (no detail data fields), no clearing method details
were imputed.
2. If the survey response for any clearing method detail field or for Return Items was "Don't
Know," the clearing method detail data were imputed for volume and value.
3. If the field was left blank for Federal Reserve Receipt Items, On-Us Deposit Items, or
Return Items, volume and value were imputed.
4. If the field was left blank for any other detail field (Clearing House, Same-Day
Settlements, Other Paid Check Volume), the volume and value were set to zero.
5. The imputation method was done as a linear proportion using PCD values for “nearest
neighbor” with next highest and next lowest actual responses for that data element. Any
imputed values were created in the same proportion of those responses as the PCD from
institution with missing data had to the range between the two nearest neighbors with
actual responses.
After imputation, the volumes and values of the clearing method data were proportionally
adjusted to equal the Total Paid Checks volume and value of their respective months. The
proportional adjustment was done for both imputed and respondent provided data.
6.2
6.2.1

Appendix B – Statistical Methodology for The Check Sample Study
CSS Sample Design

For the first phase of the sampling for this study, an approach similar to The DFI Check Study
was used, with the same three institution type strata constructed, followed by size (PCD) based
strata within each. For commercial banks, a certainty stratum was established that contained
many of the same banks in the first certainty stratum for The DFI Check Study.
The second phase of the sampling addressed the allocation of checks. Since the goal of the CSS
was to describe the universe of checks, the sample for the CSS was based on a stratified twostage design. This was a way of ensuring the same probability of selection for each check from
the universe of interest. The approach to the first stage of sampling was similar to the approach
for The DFI Check Study. The primary strata were defined on the basis of type of institution,
i.e., commercial banks (CMB), credit unions (CUS) and thrifts (THR). The next level of
stratification was carried out on the basis of size where the measure of size was the PCD value at
the highest institutional level (e.g., holding company). The end of first quarter 2001 PCD value
for each institution was used to generate a required sample number above a minimum number.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

Like The DFI Check Study, a certainty stratum was defined for the commercial banks. And like
The DFI Check Study, the sampling unit was the DFI at its highest institutional level (e.g.,
holding company), with the data collected for all the institutions owned by the sampled
companies. In contrast to The DFI Check Study, beyond the certainty stratum, the sampling
strata within type of institution were defined by selecting a minimum number of strata for
projected efficiency, then allocating the frame to those strata in such a way that the total PCD
values for each were approximately equivalent. Although this was a study of checks, minimizing
the number of institutions to be recruited to provide checks made the data collection effort more
efficient and cost effective. Additional reasons for using this approach for this stage were that it
was similar to The DFI Check Study approach, enabling comparison to The DFI Check Study
findings, and it ensured representation of the largest institutions, especially the top 100 banks.
The number of institutions and strata for each type are reported in the table below.
Table 47: Stage One Sample Allocation – DFIs Sampled per Stratum
Stratum
Top '100'
1
2
3
4
5
TOTAL

Commercial Banks
87
31
42
31
59
62
312

Type of Institution
Credit Unions

Thrifts

41
31
23
40

17
43
44

135

104

TOTAL
87
89
116
98
99
62
551

Once the stage one sample was selected, the second stage of the sample design was implemented
to allocate a required number of checks to each institution in the sample. The target number of
checks was set at 40,000 to achieve the desired precision across the types of institutions in the
study. The required number of checks was initially set based on assignment proportional to PCD
values. The approach defined the number of checks per institution, ranging from over 3,200 for
the largest bank to one check for smaller banks. The design required that a minimum number of
checks be required from each participating institution. This number was initially set at 100 to
achieve a desired total number of checks of 40,000, based on an assumed response rate of 65
percent. Requesting 100 checks was done in part to ensure institutions took the effort seriously
enough to devote the necessary resources to the data collection effort. Based on feedback
obtained during pretesting of the study materials, the minimum number per institution was
reduced to 90 checks, thereby lowering the target to 36,000 checks. Those institutions for which
the PCD allocation led to more than 90 checks became a certainty stratum for each type of
institution. The remaining 90-check institutions could then be sampled from within the
predefined size strata to fulfill the required contribution for that stratum.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

85
Table 48: Stage Two Sample Allocation – Checks Desired per Stratum (Design)
Stratum
Top '100'
1
2
3
4
5
TOTAL

Commercial Banks
12,451
1,523
1,952
1,475
2,808
2,951
23,160

Type of Institution
Credit Unions
2,080
1,475
1,095
1,904
6,554

Thrifts
2,146
2,046
2,094
6,286

TOTAL
12,451
5,749
5,473
4,664
4,712
2,951
36,000

CSS Sample Weighting

Data collected from The Check Sample Study (CSS) were inflated to the universe level using
weights designed to (1) compensate for unequal selection probabilities; and (2) adjust for nonresponding depository financial institutions (DFIs). As discussed in Section 6.2.1, the CSS
employed a two-stage sample design to select the sample of checks. The DFIs were selected at
the first stage of sampling, and the checks from the sampled DFIs were sampled in the field at
the second stage. The DFIs in the certainty strata were all selected for the study. The sample of
DFIs from the non-certainty strata was selected with probability proportional to size (PPS)
sampling from within strata. The sample of checks was selected independently from each
sampled (respondent) DFI using systematic sampling procedure. Two weights were constructed
corresponding to the two stages of sampling, i.e., a DFI weight to represent the non-sampled and
non-respondent DFIs and a check weight to represent the non-sampled checks within the
respondent DFIs. The final CSS weight was obtained by multiplying the two weights, i.e., the
DFI and check weights.
The size measure for the PPS sampling was the PCD value. The selection probabilities will vary
within strata due to PPS sampling except for the certainty strata. The base weight for a sampled
DFI was defined to be the reciprocal of selection probability of the DFI. If π hi is the selection
probability for the DFI labeled i in stratum h then the base weight (or design weight) assigned to
the corresponding DFI is defined as
Whi = 1 π hi .

The base weights for the DFIs in the certainty strata were equal to unity. The base weights were
adjusted to account for the non-responding DFIs. The nonresponse adjustment factor was
applied within each stratum. The nonresponse adjustment factor was defined as the ratio of the
number of DFIs sampled from the stratum and the number responding from that stratum.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

This weight adjustment was applied to increase the weights of the sampled DFIs for which data
were collected. As discussed above, the nonresponse weight adjustment was applied at the
stratum level. The weighted adjusted for nonresponse were simply the product of the base
weights and the nonresponse adjustment factor for the stratum. Suppose that nh DFIs were
sampled from stratum h and data were collected for rh DFIs only. Let Ah( nr ) denote the weight
adjustment due to the nonresponding DFIs within stratum h, then
Ah( nr ) =

In order to compensate for the nonresponse, the base weight for the sampled DFI i in stratum h,
i.e., Whi , for which checks were sampled was multiplied by the nonresponse adjustment factor
Ah( nr ) given above. The design weight adjusted for nonresponse then becomes

Whi( bank ) = Ah( nr )Whi .

We also applied ratio adjustments to the strata that were originally designed as certainty strata
but weights had to be applied to these DFIs due to nonresponse.
At the second stage of sampling, the checks were sampled with an equal probability systematic
sampling procedure from within responding DFIs. Let M hi be the total number of checks
deposited annually for the DFI labeled i in stratum h and mhi be the number of checks that were
sampled. Then the corresponding check weight is given by
Whi( check ) =

M hi

mhi

It should be noted that the same check weight is applied to all sampled checks from a given DFI.
The final weight for the estimation of various categories of check volume and value was
obtained by multiplying the above two weights, i.e., the DFI weight and the check weight within
the DFI. Thus, the final weight Whi( final ) was defined as
Whi( final ) = Whi( bank ) * Whi( check ) .

The super-script (final) denotes that it is the final weight for analysis purposes and the final
weight Whi( final ) was the analysis weight for all sampled checks from the responding DFI i in
stratum h.
6.2.3

CSS Estimates

In this section we discuss the estimation procedure for various categories of payer, payee and
purpose or cross-classifications of these categories. We discuss the procedure for a category or
domain denoted by d where domain can be defined from any cross-classification of the observed
categories of payer, payee and purpose. For example, a domain can be defined as counterparty
RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

equal to "consumer to consumer (C2C)" and purpose equal to "casual." The domain estimates
can be constructed by defining an indicator variable to flag the presence or absence of the
characteristics (or domain) of interest as follows.
= 1; if (hij) ∈ d
δ hij 
.
= 0; otherwise
The symbol (hij) denotes the check j sampled from DFI i in the design stratum h. The estimated
number of checks for the domain d was obtained as
VOL( d ) = ∑∑Whi( final ) ∑ δ hij .
h

Similarly, the value of those checks in the domain denoted by d can be estimated as

VAL( d ) = ∑∑Whi( final ) ∑ δ hij yhij ,
h

where yhij is the dollar amount on the sampled check j from the DFI i in stratum h.
6.2.4

Random Sampling in the Field

In an effort to maintain as much methodological control as possible over the sampling process,
while at the same time sampling in the most appropriate and efficient way for each institution,
the survey research team developed the Sampling Parameters Request Form (Appendix E). The
form was a pre-survey instrument or screener that allowed each DFI to describe the environment
in which checks would be sampled. An institution could indicate, for example, its number of
processing facilities; the average monthly volume of checks captured at each facility; whether its
check archival system assigns a unique trace number or sequence number to each item in
archive; and whether checks are archived on microfilm, digital image media or a combination of
the two.
This information allowed the survey research team to design a customized set of sampling
instructions for each DFI that completed the Sampling Parameters Request Form – instructions

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

that ensured the most random and representative sample of checks. If an institution had multiple
processing facilities, for example, the survey research team specified exactly how many checks
should be sampled from each of the institution's processing facilities, from what dates and from
exactly where in the sequence of each day's check processing volume.44
6.2.4.2 Master List of Random Checks

This information was conveyed to institutions that completed the screener through a Master List
of Random Checks, which accompanied each institution's sampling instructions. The Master List
included for each item to be sampled the date on which it was processed and a specific Random
Check Number. Dates were chosen at random from the 252 eligible processing days in the May
1, 2000 to April 30, 2001 survey period. The Random Check Number for a given date was
chosen at random between 1 and the institution's average daily volume of checks processed.
The average daily volume was calculated using volume data provided via the screener. For
institutions that commingle deposited checks with inclearings, the upper bound on this Random
Check Number was the average daily prime pass volume – essentially, the combined volume of
both inclearings and deposits. This ensured that each deposited check had an equal probability of
being selected from the commingled archive, regardless of whether it fell at the end of a batch of
inclearings. For institutions that archive deposited checks separately from inclearings, the upper
bound on the Random Check Number was simply the average daily deposit volume.
The Random Check Number meant different things to different institutions, depending on their
processing environment and archival practices. We could not expect, for example, an institution
with 5 processing facilities and 4-5 sorter devices per site – all working in parallel – to choose
the 5,439th item on a given date. How would the respondent know which item to select from a
central archive? As a practical matter the notion of a Random Check Number worked well for
smaller institutions but was overly simplistic for many of the larger institutions in the survey.
Therefore, Global Concepts developed a process of randomizing each variable in a DFI's archival
environment to best accommodate its particular sampling requirements. This process was highly
customized and labor-intensive, but it paid off in the randomness of the sample.
6.2.4.3 Master List of Random Sequence Numbers

The primary method of randomizing the sample was through the use of Random Sequence
Numbers. Many institutions perform check photo copy retrieval based on an indexing scheme of
sequence numbers or trace numbers. A Random Check Number was meaningless to many
institutions that require a sequence number to retrieve film. For these institutions the research
team provided instructions for how to create a Master List of Random Sequence Numbers. In
practice, this was the most cumbersome aspect of the sampling process and often was
disregarded in favor of less onerous methods of sampling checks at random. Nevertheless, many
institutions relied on a Master List of Random Sequence Numbers as the basis for photo retrieval
of sample checks.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

Often the Master List was generated manually by Global Concepts. Global Concepts worked
closely with dozens of financial institutions to document exactly how their individual sequence
numbering schemes were organized. By understanding the variables that constitute a sequence
number (e.g., site number, sorter device number, date, etc.) as well as the average daily range of
each of those variables, Global Concepts was able to produce a customized Master List of
Random Sequence Numbers for each institution that requested one. These lists included
numerous Alternate Random Sequence Numbers in case the principal Random Sequence
Number did not point to an actual item for the given date. In that event, the Alternate Random
Sequence Number was used, and so on.
6.2.4.4 Random Sampling Wizard

Some institutions were made self-sufficient in the creation of a Master List of Random Sequence
Numbers with the help of a software tool Global Concepts developed specifically for that
purpose. The Random Sampling Wizard was sent along with the sampling instructions to all
institutions who indicated on the screener that their photo retrieval process requires the use of
sequence numbers. The Wizard was essentially a Microsoft Excel file with Visual Basic macros
that walked the respondent through a series of questions about the institution's sequence
numbering scheme. The Wizard asked, for example, which variables constitute a sequence
number, their sequential order in the sequence number format, the average daily range of each
variable (e.g., how many sorter device operate per day), the appropriate format of each variable
(e.g., how many digits, the proper date format), and so on. The Wizard then randomized each
variable in the sequence number and produced a Master List of Random Sequence Numbers
complete with random dates and several Alternate Random Sequence Numbers.
The Wizard ultimately proved to be overly simplistic for the diversity of sequence numbering
schemes in practice in the industry. A primary shortcoming was the Wizard's inability to
accommodate sequential gaps in the component variables of a sequence number. Many
institutions with multiple processing sites include a sorter device number field in their sequence
numbers, and these sorter numbers often include large gaps in their sequence. For example, the
devices at Smallville may be numbered 01 through 07, and the devices at Metropolis numbered
20 through 25 with neither site accounting for numbers 08 through 19. In other cases an
institution may have sorters 25 through 27, sorters 34 through 36 and sorter 56 all in the same
processing facility. These situations pushed accurate list generation beyond the capabilities of the
Random Sampling Wizard, which would attempt to randomize the sorter device number between
the lowest and highest numbers provided without consideration for sequential gaps. In short, it
was prone to producing sequence numbers that did not exist. As such, Global Concepts generally
performed custom development of sequence number lists for these institutions. In general,
Master List production was a highly customized process, whether the institution chose a Master
List of Random Sequence Numbers or otherwise.
6.2.4.5 Customized Randomization Schemes

Dozens of institutions – even those who generally use sequence numbers to perform photo
retrieval – opted for another customized approach than one based on Random Sequence
Numbers. Formatted correctly or not, a Master List of Random Sequence Numbers is no
guarantee in itself that each Random Sequence Number on the list is an actual sequence number
in the archive. Many institutions preferred to avoid the process of hit-and-miss.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

Global Concepts worked with numerous institutions to create custom Master Lists that
randomized each of a series of known variables. A typical scenario involved randomization of
the site number, sorter device number, microfilm roll number and the location of the check on
the chosen roll of film. Information about each of these variables, such as the number of sorter
devices per site or the average daily number of microfilm rolls per sorter, were provided by the
respondent to Global Concepts in order to create the Master List.45 Global Concepts randomized
each variable to produce a list – not of sequence numbers – but of individually randomized
variables. The respondent would then retrieve, for example, a check from the second roll of
microfilm produced by sorter device #3 at the Smallville location on a randomly generated date.
The location of a check on a roll of microfilm was randomized using a random time variable.
Photo retrieval staff were instructed to advance the roll of microfilm for a particular number of
seconds and to choose the nearest BFD item to their stopping place. Institutions that use image
archives frequently would use a list of randomly generated dates combined with a random time
variable if the photo retrieval system allowed them to scroll through the day's volume of checks
and then stop at the specified random time.
6.2.4.6 Photo Retrieval Latitude

As noted above, there was no guarantee that a Random Check Number or Random Sequence
Number provided by Global Concepts would point to an actual deposited check. In some cases,
the Random Number would exceed the total volume processed that day. In other cases, the
Master List may have listed an item for which the DFI was not the bank of first deposit, a deposit
slip or a general ledger (GL) ticket.46 For that reason each Master List included alternate Random
Check Numbers or Sequence Numbers to increase the likelihood of a "hit" on a given date. Even
these were no guarantee. Therefore, all institutions were instructed to find the deposited check
nearest to the Random Check Number or Sequence Number specified. Generally, this meant
rewinding microfilm or advancing to the next item in an optical image archive.
6.2.4.7 DFI-Selection of Sampled Checks

Not all respondents who completed The Check Sample Study responded to the Sampling
Parameters Request Form. In total, 127 of the 149 institutions that surveyed a random sample of
checks completed the screener prior to surveying a sampled of checks. For the remaining
institutions the survey research team could not specify – without communication with these DFIs
– the proportion of checks to sample from each of their processing sites or the Random Check
Number to be retrieved. We lacked the information needed to make an accurate estimate of either
of these factors. Therefore, a more simplified Master List was provided to these institutions.
For the 22 DFIs that did not complete the Sampling Parameters Request Form, the research team
provided a randomly selected date from which to sample BFD items and specific instructions
about how to ensure the most random and representative sample of BFD items. These
instructions included clear guidelines about weighting the sample across multiple processing
45

Global Concepts used the exact sorter numbers or site numbers in the randomization routine to avoid the nonsequential difficulties that crippled the Random Sampling Wizard. While time consuming on the front end, this
proved highly effective in terms of a "hit ratio" for photo retrieval on the back end.
46
BFD items are distinguished by the presence of only one DFI endorsement.

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

sites (if applicable) in proportion to the proof-of-deposit volume processed by each of those sites.
As a practical matter many of these institutions contacted Global Concepts after receiving the
survey materials, and Global Concepts provided a customized Master List as described above.
6.3

Appendix C – The DFI Check Study Survey Instrument

(Follow this link.)
6.4

Appendix D – The Check Sample Study Survey Instrument (Answer Sheet)

(Follow this link.)
6.5

Appendix E – The Sampling Parameters Request Forum (The Screener)

(Follow this link.)

RETAIL PAYMENTS RESEARCH PROJECT

Sponsored by the Federal Reserve System

Appendix C:
The DFI Check Study Survey Instrument

Depository Financial Institution Check Study
Survey of Paid Check Volume and Value
Survey Period: March 1, 2001 – April 30, 2001

IMPORTANT… Report paid checks only.
t

Do not include ATM, ACH or other debits.

Do not include transit items.

Include only your Inclearings and over-thecounter On-Us checks.

A project conducted by The Federal Reserve, Global Concepts and Westat

About the Study
This survey is part of the Federal Reserve's Check and Electronic Payments Research Project. The
aim of this project is to establish the total volume and value of U.S. non-cash payments, identify
emerging payment types, and provide data to help the industry and the Federal Reserve to identify
opportunities for improving the efficiency of the nation’s payments system. The data we are asking
you to collect as part of The Depository Financial Institution Check Study will be used to estimate
the total volume and value of checks paid in the United States.
Please read all instructions and the Appendix of Important Terms carefully before you start
completing the survey. To ensure consistency of responses, it is important that your institution use
the definitions of paid checks provided. If you have any questions, please call Westat toll free at
(888) 263-9854. One of the Westat staff will be happy to answer your questions or direct you to the
appropriate individual at Global Concepts or the Federal Reserve.

Key Dates and Instructions for Returning Survey
t The survey period is March 1, 2001 – April 30, 2001.
t Please fill out all three pages of the survey, including the questionnaires for:

—

Paid Check Volume and Value (dollar value rounded to nearest $1,000s)

—

Outgoing Return Checks Volume and Value

—

Active Routing Transit Numbers

t Respond to every item on the surveys of Check Volume and Value. If an item does not

apply, or if its volume or value is zero – please enter a zero in the space provided.
t Respond to all items at the highest institutional level (i.e. include volume from all

subsidiary institutions and processing sites).
t Before May 18, photocopy your completed Survey Forms so that you have a backup on

file. Keep your backup copies on hand until August 1, 2001.
t Please send us your completed Survey Forms by Friday, May 18, 2001. You may mail

them, fax them, or enter your data online.
Option 1:
Mail to… DFI Check Study

Option 2:

Option 3:

Fax to… (888) 783-1782

Visit…

http://www.checkstudy.com

c/o Westat
1650 Research Blvd.
Rockville, MD 20850

Thank
Thank you…for
you
your time and effort on behalf of the Federal Reserve. We understand that your participation in
this study involves a time commitment for some of your staff. We appreciate your willingness to assist us in
gathering this data and creating valuable information for the entire financial services industry.

Institution ID: *****************

Survey of Paid Check Volume/Value: March 1 - April 30, 2001
Note: Total Paid Checks is the sum of all items (a-e) below it. Also, before starting, please review the Appendix of
Important Terms beginning on page 6. Please include any Merged Volume. Round dollar amounts to the nearest
thousand. If the volume and value of an item can not be determined, please check, “Don’t Know.”

MONTH 1:
1.

1a.

1b.

1c.

1d.

1e.

2a.

2b.

2c.

2d.

2e.

Volume of paid checks

Value of paid checks

Don't
Know

TOTAL PAID CHECKS (MARCH) — Total
received by your institution March 1 - 31, 2001.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Federal Reserve Receipt Items — Total
received from the Federal Reserve presented by
the Federal Reserve. Not Same Day Settlement.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

On-Us Deposit Items — Total received by
your institution for which you were the BFD.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Clearing house / Local Exchange Items —
Total received from clearing houses or other local |__||__|__|, |__|__|__|, |__|__|__|
exchanges. Not Same Day Settlement.

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Same Day Settlement Receipt Items —
Total received under Same Day Settlement rules.
Not Federal Reserve or Clearing House Items.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Other Paid Check Volume — Include any
volume that you did not or cannot allocate above.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

MONTH 2:
2.

MARCH 1 – 31, 2001

April 1 – 30, 2001

Volume of paid checks

Value of paid checks

Don't
Know

TOTAL PAID CHECKS (APRIL) — Total
received by your institution April 1 - 30, 2001.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Federal Reserve Receipt Items — Total
received from the Federal Reserve presented by
the Federal Reserve. Not Same Day Settlement.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

On-Us Deposit Items — Total received by
your institution for which you were the BFD.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Clearing house / Local Exchange Items —
Total received from clearing houses or other local |__||__|__|, |__|__|__|, |__|__|__|
exchanges. Not Same Day Settlement.

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Same Day Settlement Receipt Items —
Total received under Same Day Settlement rules.
Not Federal Reserve or Clearing House Items.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Other Paid Check Volume — Include any
volume that you did not or cannot allocate above.

|__||__|__|, |__|__|__|, |__|__|__|

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Institution ID: *****************

Survey of Return Checks (Outgoing): March 1 - April 30, 2001
Please indicate the number of outgoing returns. Please provide a total for checks only sent by your institution
during each month of the survey period. Do not include electronic returns. Please include any Merged Volume.
Round dollar amounts to the nearest thousand. If the volume and value of your returned checks can not be
determined, please check, “Don’t Know.”

MONTH 1:
3.

Return Checks (Outgoing)
Total checks returned unpaid during the survey
period.

MONTH 2:
4.

MARCH 1 – 31, 2001

April 1 – 30, 2001

Return Checks (Outgoing)
Total checks returned unpaid during the survey
period.

Volume of checks
Returned

|__||__|__|, |__|__|__|, |__|__|__|

Value of checks
returned

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

Volume of checks
Returned

|__||__|__|, |__|__|__|, |__|__|__|

Value of checks
returned

$ |__|__|__|, |__||__|__|, |__|__|__|, 000.

You have completed the volume/value sections of the survey.

Please complete the Survey of Routing Transit Numbers on
page 4.

To avoid double counting of items, it is important that we
understand what RTN volume and value data are included in your
totals.

If you prefer to submit a pre-formated report that shows your
RTN’s, please just attach it to the RTN Survey form.

Institution ID: *****************

Don't
Know

Survey of Active Routing Transit Numbers
Please write in all of the active 9-digit Routing Transit Numbers (RT’s) owned by your institution that correspond
to the paid check volume/value reported in the previous two tables. Also indicate which, if any, RT's were acquired
through a merger or acquisition during March 1, 2001 – April 30, 2001, and when they were acquired.
If necessary, please photocopy and fill out additional copies of this page. If you prefer, you are welcome to
however, please be sure to indicate on your documents which
RT’s were acquired during the survey period and when they were acquired

present a list of RT's on pre-existing reports;

9-digit routing transit number

Institution ID: *****************

Acquired
during survey
period?

If "yes,"
Date RT
acquired

9-digit routing transit number

Acquired
during survey
period?

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

q
q

Yes
No

If "yes,"
Date RT
acquired

Your Comments…
Thank you for completing the survey.

Please use the space below to provide any comments or suggestions that
you feel might be useful for us to better understand your response or to improve subsequent versions of this survey.
For example, you might comment about the effort required to report the various data elements, the time required to
respond, the clarity of the definitions, the design of the survey instrument, etc.

Institution ID: *****************

APPENDIX
Important Terms You Need to Know
Please carefully read the definitions of the survey terms that we use in our materials. Although some of the terms
may be familiar to you, their definitions may differ slightly from those commonly used within your institution. To
make sure that our results are comparable, it is important that all financial institutions understand each term to mean
the same thing and report data accordingly. So, please take the time to review the General Terminology and Survey
Data Elements that follow. If you have any questions, please feel free to call us.
General terminology
Check

– A negotiable instrument drawn on a financial institution. For this study, please follow our
guidelines:
Checks do not include…

Checks include…

t Checks written by individuals, business or
government entities

t Deposit slips
t Rejected items (i.e., checks)

t Traveler's checks

t General ledger tickets

t Money orders

t Other non-check documents, such

t Cashier's checks

as payment coupons

t Teller's checks
t Payable through drafts
t Truncated checks (i.e., ECP file items)

Paid Check

— A check for which your institution is the payor bank as defined by Regulation CC.

t Do not include checks that you receive as a “pass through correspondent” for which another

institution is actually the payor bank.

Electronic Check Presentment (ECP)

— ECP items are paid checks that you — the paying bank —
receive in an electronic file that may be used for posting. ECP items are considered received
when the electronic file arrives. ECP includes two types of volume:
t Truncation: If your institution receives paid checks via ECP and no corresponding paper

checks follow, be sure to count this volume in your results.
t Paper to Follow: If your institution receives paid checks via ECP and the corresponding

paper checks follow at a later date, do not double count these items — once in the electronic
file, and again when the physical items arrive.

Institution ID: *****************

General Terminology (continued…)

Merged Volume — The total combined volume
merge during the survey period.

of your institution and any institution with which you

Note:

If at any point during the survey period another institution merges with your institution,
please report your merged volume for the entire survey period – as if the merger had occurred the
day before the survey period began.

Survey Period
2001.

— Include all paid checks received during the survey period: March 1, 2001 – April 30,

Received Checks

— An item is received when it is:

t Presented by another depository financial institution, either directly or through a clearing
t
t
t
t

arrangement (e.g., Clearing House)
Presented by the Federal Reserve Bank
Deposited
Cashed
Received as payment (e.g. loan payment)

Since the date that an item is received may be different from the date the work is
processed:

t Include volume you receive during the evening of April 30, 2001 that you do not process until

the following day (after the survey period) — for example, a batch of branch deposits received
on April 30, 2001 but processed on May 1, 2001.
t Do not include volume you receive on February 28, 2001 (the day before the survey period),

that you process the morning of March 1, 2001 — for example, late night presentments from
the Federal Reserve on February 28, 2001.

Institution ID: *****************

Survey Data Elements

1, 2.

Total Paid Checks

1,2a.

Federal Reserve Receipt Items

— All paid checks received by your institution during the survey period. This total
should be the sum of all data elements defined below.
— All paid checks received from the Federal Reserve Bank that are
presented by the Federal Reserve Bank during the survey period.
t

Do not include Same Day Settlement Volume if the Federal Reserve is your
designated presentment point for Same Day Settlement. This includes Same Day
Settlement volume that is:

— Delivered by, but not processed by, the Federal Reserve.
— Processed, truncated, and delivered by the Federal Reserve as an ECP file.

Same Day Settlement volume received from the Federal Reserve should be reported in the
Same Day Settlement category (below).
t

Include

truncated items that are not Same Day Settlement items if the Federal Reserve
truncates checks presented to your institution as part of an ECP service.

1,2b.

On-Us Deposits

1,2c.

Clearing Houses and Other Local Exchange Items

— All paid checks received during the survey period through your institution's
branches, ATMs, lockbox operations, cash vault, payment processing centers, etc. This is
volume for which your institution is both the payor bank and the bank of first deposit (BFD).

— All paid checks received during the survey
period from a clearing house or other local exchange relationship.

Do not include Same Day Settlement

volume in your clearing house totals if a clearing
house is your designated presentment point for Same Day Settlement or acts as the
settlement agent for Same Day Settlement. This volume should be reported in the Same
Day Settlement category (below).

1,2d.

Same Day Settlement

1,2e.

Other Paid Check Volume

— All paid checks received under the rules of Same Day Settlement. This
includes items received through the Federal Reserve or a clearing house acting as your
designated presentment point but settled directly with the presenting bank.
— A sub-total of paid check volume. All paid checks received during the
survey period that satisfy one or both of the criteria below:

It does not meet any definition

of items (1,2a-d) above – such as volume received from

a correspondent processor.
t

3,4.

It cannot be allocated

accurately to another category – such as Clearing House, Same
Day Settlement or On-Us volume that you "don't know."

Return Checks (Outgoing) —

Checks drawn on your institution that you return unpaid during the

survey period.

Institution ID: *****************

Appendix D:
The Check Sample Study Survey Instrument
(Answer Sheet)

Coding a Check / Answer Sheet
1. Are any of these words on the front of the check?
m Cashier's Check or Certified Check
m Money Order or Postal Money Order
m None of the above

Payee (paid by the check)
6. Does the Payee Line or the check itself include an
address for the Payee?

¬
-

Payer (wrote the check)
2. Does the Payer name or address have any of these?
(Check all that apply)

7. Does the Payee name (or address, if present) have
any of these? (Check all that apply)
m One or more persons' full names (John Smith, John and

m One or more persons' full names (John Smith, John and

Mary Smith, John Smith, Inc.)

m Cash
m Inc., LLC, LTD, Co., NA, Corp., Corporation, Trust,

m Inc., LLC, LTD, Co., NA, Corp., Corporation, Trust,
Trustee, Company, Services, .com, Association, PC

m Bank, Insurance
m Initials of Business or Association (e.g. NAACP, AT&T)
m State of, City of, County of, Town of, Township of, Bureau

Trustee Company, Services, .com, Association, PC

m Bank, Insurance
m Initials of Business or Association (e.g. NAACP, AT&T)
m State of, City of, County of, Town of, Township of, Bureau

of, Municipality

m State Treasury, State Treasurer, County Treasurer,

of, Municipality

County Commissioner, County Controller

m IRS, Internal Revenue Service, State Tax, County Tax,

m Port Authority, Water Authority, Power Authority, Transit

Tax Commissioner, Tax Collector

Authority, Department of

m
m
m
m
m

¬
®
¯
°
±

Authority, Department of

School, High School, Elementary, University, College

m
m
m
m
m
m

Mail code (e.g., MC-648)
Accounts Payable, Acct. Payable
NO -- None of the above

Consumer
Government
Business
Not Consumer – either business or government
Not Government – either business or consumer
Cannot determine

4. Your primary reason for categorizing the Payer:

¬
®
¯
°
±
²
³

m Port Authority, Water Authority, Power Authority, Transit

Apartment number (apt. #) NOT Suite # or Building #

3. Based on the Payer name and address and the
characteristics of the check, can you definitively
categorize the Payer as any of these?

Consumer name
Familiar with business name

Yes
No

School, High School, Elementary, University, College
Dr., Doctor, MD, DDS, DVM, PC, Specialist, –ologist
Apartment number (apt. #) NOT Suite # or Building #
Mail code (e.g., MC-648)
Accounts Receivable, Acct. Receivable
NO -- None of the above

on the front…
8. Date of the check:

9. Dollar amount of the check:
$

10. Is the dollar amount in the Courtesy Amount Field
handwritten?

¬
-

Yes
No – it is machine printed

11. The 9-digit transit routing number:

Familiar government organization
Familiar, but unclear whether business or government
Not familiar, but clearly a business name
Not familiar, but clearly a government organization
Not familiar, but clearly business or government
Cannot categorize

5. Payer's ZIP code:

12. Is this a large format check? (hint: the symbol
II
will appear somewhere to the left of the transit
routing number.)

¬
-

Yes
No

13. Is the memo line used?

¬
-

Yes
No

If it is used, write the first 12 characters:

14. Is the signature on the face of the check handwritten or a facsimile of a hand-written signature?

¬
®

Yes – Hand-written or facsimile
No – Name in printed type face or "No Signature
Required"
There is no signature, hand-written or otherwise.

15. Are any of these items handwritten on the check?
(Check all that apply)
m DL, driver's license, license
m Handwritten state initials (GA, CA, MI, etc.) followed by or
preceded by a number

m Account, (e.g. acct #) followed or preceded by a number
m Phone number handwritten or circled on face of check
m Birth date written on check (Note: Date will be 1990 or
earlier.)

18. On the back, is the Payee endorsement
perpendicular or parallel to the writing on the front of
the check? (Refer to Answer Sheet Guide example.)

¬
®

Perpendicular
Parallel
Cannot find Payee endorsement

19. Do the words "Absent(ee)” or “Absent Endorsed”
appear anywhere on the back of the check?

¬
-

Yes
No

Categorizing Payee
20. Based on the Payee name/address and
endorsement, can you definitively categorize the
Payee as any of these?

¬
®
¯
°
±

Consumer
Government
Business
Not Consumer – either business or government
Not Government – either business or consumer
Cannot determine

m Stamped form (generally on the back of the check) that is
filled in with handwritten characters

NO -- None of the above

on the back…
16. Are any of these words in the Payee endorsement?
(Check all that apply)
m Dollar Amount, Amount, $
m Store, Store #, register #, terminal #, branch #, location #,
DL, D/L, cashback

m Inc., LLC, LTD, Co., NA, Corp., Corporation, Company,
Services, .com, Association, Trust

m Bank, Insurance
m Initials of a Business or Association (NAACP, AT&T)
m State of, City of, County of, Town of, Township of, Bureau
of, Municipality

m IRS, Internal Revenue Service, State Tax, County Tax,
Tax Commissioner, Tax Collector

m Port Authority, Water Authority, Power Authority, Transit
Authority, Department of

m School, High School, Elementary, University, College
m Dr., Doctor, M.D., DDS, DVM, PC, Specialist,
–
ologist

m NO -- None of the above
17. Is the Payee endorsement handwritten?

¬ Yes – Handwritten
- No – It's stamped / machine-printed
® Cannot find Payee endorsement

21. Your primary reason for categorizing the Payee:

¬
®
¯
°
±
²
³

Consumer name
Familiar with business name
Familiar government organization
Familiar, but unclear whether business or government
Not familiar, but clearly a business name
Not familiar, but clearly a government organization
Not familiar, but clearly business or government
Cannot categorize

22. If Payee is business or government, mark which
type:

¬ Power, gas, phone, cable or internet service provider
- Bank or credit card company or insurance company
® Supermarket or Drugstore
¯ Convenience store
° Retail Store, retail service shop, or cataloger
± Restaurant, bar, diner, fast food, etc.
² Subscription, membership organization, club, etc.
³ Charitable organization, church
´ Medical (e.g., hospital, doctor’s office, etc.)
Other business or government (not individual consumer)

µ NOT a business or government

Appendix E:
The Sampling Parameters Request Form
(Screener)

Check Sample Study
SAMPLING PARAMETERS REQUEST FORM

A Preliminary Screener for the Check Sample Study

A project conducted by The Federal Reserve, Global Concepts and Westat.

If you have any questions about what you need to do, please call us toll free at (888) 458-8608.
We’ll be happy to answer your questions.

Contents
Background........................................................................................................................... 1
When is Your Response Due................................................................................................ 2
Definitions of Important Terms............................................................................................. 2
Answer Sheet – Check Processing Profile Information ...................................................... 3

Background
The Federal Reserve is conducting the Check Sample Study to estimate broad categories of US check payments,
such as the number of checks being written by consumers and businesses and the purpose of these payments, such
as bill payments or point-of-sale purchases.
The Check Sample Study requires that you survey a representative random sample of ________ deposited checks.
You will be asked to select these items at random from your microfilm or image archives and record on an answer
sheet information about each check's characteristics. The answer sheets will be sent in early May as part of the
Data Collect Guide.
The purpose of this Sampling Parameters Request Form is to make the task of retrieving sample checks as easy as
possible. To further simplify the process we encourage you, at your earliest opportunity, to notify your photo
retrieval staff that they will be helping to select a random sample of deposit items later this Spring. Although the
study has a few special requirements – e.g. retrieving only deposited checks – the retrieval process has been
designed to complement their existing procedures.
Based on your response to the Sampling Parameters Request Form, we will send you the Data Collection Guide,
complete with instructions for retrieving and surveying a representative random sample of deposited checks. The
sampling and surveying should be performed in early May.
Thank you for your time and effort on behalf of the Federal Reserve. We understand that your participation in this
study involves a time commitment for some of your staff. We appreciate your willingness to assist us in gathering
this data and creating valuable information for the entire financial services industry. If you have any questions
about what you need to do, please call toll free: (888) 458-8608. We’ll be happy to answer your questions.

Sampling Parameters Request Form

Institution ID:

______________

When is Your Response Due
Here’s what you need to know about completing this preliminary Request Form on time:
t By April 30 please complete, photocopy, and return page 3-4 of the Request Form along with any

additional attachments. Keep your photocopy on file as a backup until August.
t You may either mail or fax your completed Request Form to us. Alternatively, you may use our secure
web site to enter data securely online: http://www.checkstudy.com.
Mail Request Form Response To:

Or Fax Request Form Response To:
(888) 783-1782

Federal Reserve Check Study
c/o Westat
1650 Research Blvd.
Rockville, MD 20850

Definitions of Important Terms
Deposited Check Volume –

All deposited checks received by your institution through any of
various methods (e.g. over-the-counter branch deposits, ATM deposits, lockbox deposits,
cash vault deposits, etc.). This includes POD (proof-of-deposit) volume and PED (preencoded deposit) volume. It does not include inclearings of any kind.

Inclearings – Checks

received from other institutions. This includes checks drawn on your
institution or an institution for which you provide processing services.

Prime Pass Volume –

All checks processed by your institution or by a processor on your
behalf. This includes Deposited Check Volume as well as inclearings and other items.

Sequence Numbers – Unique

identification numbers, typically referred to as sequence numbers,
document identification numbers, or trace numbers that are assigned to each deposited check as it
is processed. Sequence numbers are used as a reference when retrieving checks for research and
adjustments purposes.

Survey Period – This

study will survey checks deposited during May 1, 2000 – April, 2001.

Sampling Parameters Request Form

Institution ID:

______________

Answer Sheet – Check Processing Profile Information
1.

Please estimate A) your institution's average monthly volume (i.e. number) of deposited checks and
B) your average monthly prime pass volume (including deposits, inclearings and other items).
A) Avg. Deposited Check Vol. per Month:

|__|__|__|, |__|__|__|, |__|__|__|

B) Avg. Total Prime Pass Vol. per Month:

|__|__|__|, |__|__|__|, |__|__|__|

Are checks deposited with your institution processed in such a way that assigns a unique
identification number (e.g. sequence number, document identification number, trace number, etc.)
that uniquely identifies each deposited check for lookup purposes?

Yes

Is there a unique storage practice (e.g. range of "sequence numbers," separate location/media, etc.)
that distinguishes the photo records of deposits from other volume (i.e. inclearings or other items)?

Yes

By April 30, 2001 on which media will you have a photo record of all checks deposited with your
institution (both on-us and transit items) during the survey period (May 1, 2000 – April 30, 2001)?

q Microfilm

q A combination of media – i.e. no single medium

q Digital Image Media (Optical or Magnetic)

q Neither Microfilm nor Digital Images

Indicate if you do not have in-house photo retrieval capabilities for any of the following?

q Deposits – On-Us (i.e. deposits drawn on your institution.)
q Deposits – Transit (i.e. deposits drawn on another financial institution.)
q Inclearings and other non-deposit items
6.

At how many sites operated by your institution (including branches) do you capture either microfilm
or digital images of deposited checks?

0 – "All deposits processed by a third-party processor."

1 – "All volume in Question 1-A captured centrally."

2 or more – Please indicate how many: ________

– Stop. Form complete.

– Continue.

On a the following page (or by attaching your own existing reports), please…
A) List the location of each site operated by your institution (including branches) where check deposits

are captured as either microfilm or digital images.
B) For each site, indicate the average monthly volume of deposited checks captured and prime pass
volume (including deposits) captured at that site.
Sampling Parameters Request Form

Institution ID:

______________

CHECK PROCESSING SITES AND AVERAGE MONTHLY VOLUMES:
Note: If you do not outsource any check processing, the sum of all site volumes below should equal the total
volumes you provided in Question 1 above.
Processing Site (or Branch)

Monthly Deposit Volume

Monthly Prime Pass Volume

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.

Sampling Parameters Request Form

Institution ID:

______________

Full text of Federal Reserve Payments Study : Retail Payments Research Project : A Snapshot of the U.S. Payments Landscape

FRASER