U.S. Postal Service: New Focus on Improving Service Quality and Customer
Satisfaction (Chapter Report, 12/20/95, GAO/GGD-96-30).

Above all, Postal Service customers want and expect prompt, reliable
mail delivery. When dissatisfied with traditional mail services,
customers increasingly turn to alternatives, such as electronic
communication or other suppliers. According to the Postal Service, on
the basis of current customer satisfaction levels and if customers could
use another service at the same price, more than 40 percent of the
residential customer market could be vulnerable to competition. This
report examines the Postal Service's efforts to measure, report, and
improve customer satisfaction. GAO discusses (1) the extent to which the
Service distributes customer satisfaction data for use internally and by
Congress; (2) whether the Service can improve the distribution of that
data; (3) what steps it is taking to use customer satisfaction and other
performance data to improve customer satisfaction by improving customer
service; and (4) any additional steps it could take to improve customer
satisfaction.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  GGD-96-30
     TITLE:  U.S. Postal Service: New Focus on Improving Service Quality 
             and Customer Satisfaction
      DATE:  12/20/95
   SUBJECT:  Customer service
             Mail delivery problems
             Surveys
             Mail transportation operations
             Work measurement standards
             Information analysis operations
             Proprietary data
             Personnel evaluation
             Oversight by Congress
             Statistical methods
IDENTIFIER:  USPS CustomerPerfect Initiative
             USPS Customer Satisfaction Index
             USPS Mystery Caller Program
             USPS External First-Class Measurement System
             USPS Quality First Initiative
             Dept. of Commerce Malcolm Baldrige Quality Improvement Award
             USPS Easy Stamp Program
             USPS Box Call Project
             USPS Innovations Network
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

December 1995

U.S.  POSTAL SERVICE - NEW FOCUS
ON IMPROVING SERVICE QUALITY AND
CUSTOMER SATISFACTION

GAO/GGD-96-30

U.S.  Postal Service

(222016)


Abbreviations
=============================================================== ABBREV

  APWU - American Postal Workers Union
  BCSI - Business Customer Satisfaction Index
  CFS - Computerized Forwarding Site
  CSI - Customer Satisfaction Index
  EAS - Executive and Administrative Schedule
  EOS - Employee Opinion Survey
  EXFC - External First-Class
  FOIA - Freedom of Information Act
  NPR - National Performance Review
  PCES - Postal Career Executive Service
  SET - Striving for Excellence Together
  ZIP - Zone Improvement Plan

Letter
=============================================================== LETTER


B-249787

December 20, 1995

The Honorable John M.  McHugh
Chairman, Subcommittee on the Postal Service
Committee on Government Reform and Oversight
House of Representatives

The Honorable Gary A.  Condit
House of Representatives

This report responds to your request that GAO report on the Postal
Service's efforts to measure, report, and improve customer
satisfaction. 

The report contains recommendations to the Postmaster General to
improve the dissemination and use of customer satisfaction and other
performance measurement data.  Among other recommendations, the
report recommends that the Postal Service consult with appropriate
congressional oversight Committees to determine what business and
residential customer satisfaction data and what other performance
data should be regularly provided to Congress for its use. 

We are sending copies of this report to other congressional
Committees, the Postmaster General, and other interested parties. 

The major contributors to this report are listed in appendix VI. 
Please contact me on (202) 512-8387 if you or your staff have any
questions concerning the report. 

J.  William Gadsby
Director, Government Business
 Operations Issues


EXECUTIVE SUMMARY
============================================================ Chapter 0


   PURPOSE
---------------------------------------------------------- Chapter 0:1

Above all, Postal Service customers want and expect prompt, reliable
mail delivery.  When dissatisfied with traditional mail services,
customers increasingly seek and find alternatives, such as electronic
communication or other suppliers.  According to the Postal Service,
based on current customer satisfaction levels and if customers could
use another service at the same price, more than 40 percent of the
residential customer market would be vulnerable to competition. 

The Subcommittee on the Postal Service, House Committee on Government
Reform and Oversight, and its predecessor requested that GAO report
on the Postal Service's efforts to measure, report, and improve
customer satisfaction.  GAO's objectives were to determine (1) to
what extent the Service distributes customer satisfaction data for
use internally and by Congress; (2) whether the Service can improve
the distribution of that data; (3) what steps it is taking to use
customer satisfaction and other performance data to improve customer
satisfaction by improving customer service; and (4) what additional
steps, if any, it could take to improve customer satisfaction. 


   BACKGROUND
---------------------------------------------------------- Chapter 0:2

The Postal Service is in a unique position of functioning as both an
independent executive branch establishment accountable to Congress
and a businesslike entity competing with technology and private
companies to deliver mail.  Therefore, the Service is necessarily
concerned with accomplishing its mandate for universal mail service
while also advancing and protecting its competitive interests. 

The Service recognizes that both the competition it faces and its
public service responsibilities dictate that it measure and
continuously improve the quality of its services.  The Service began
its current systems of measuring residential customer satisfaction in
1991 and business mailer satisfaction in 1993.  The Service also
measures on-time delivery performance for First-Class mail, and those
results are projectable for the entire class.  It measures on-time
delivery of second-class and third-class mail for some customers, but
the results of these measures are not projectable to the entire
classes. 

The Service's Office of Consumer Advocate administers these
measurement systems.  Various vice presidents at postal headquarters
and 10 postal area offices oversee 85 "performance clusters." A
cluster denotes a geographic service area and includes a customer
service district (responsible for overseeing post offices) and one or
more mail processing plants.  Managers and employees in these
clusters are to use customer satisfaction data and other measures to
continuously improve performance and, in turn, service quality. 

The Service placed renewed emphasis on customer service when
restructuring itself in 1992 and has subsequently developed new
policies, standards, and systems to focus greater attention
throughout the Service on improving customer satisfaction.  Congress,
too, has continued to watch over the Service's performance in
providing mail service to all communities. 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

The Postal Service widely distributes residential customer
satisfaction data internally for use in improving customer service. 
The Service has shared little of that data with Congress and in
recent years reduced the amount of residential customer satisfaction
and other performance data provided in required comprehensive reports
filed annually with Congress. 

Although the Service and Congress have found residential customer
satisfaction data to be useful, those data do not tell the whole
story.  Business customers, not residential customers, generate most
(about 90 percent) of the Service's mail volume and revenue, and the
Service faces strong competition in serving these business customers. 
Because of concerns about hurting its competitive interests, the
Service has not disseminated internally or externally any business
customer satisfaction data, which it has gathered at considerable
cost.  GAO agrees that the indiscriminate release of detailed
business customer satisfaction information to the public could harm
the Service's competitive interests.  However, GAO also believes that
the risk of releasing some data must be balanced against the
potential value to the Service and Congress of using that data, with
appropriate safeguards, to help assess and improve customer service. 

Since 1990, the Service has begun many innovative and promising
efforts to improve customer satisfaction by improving the quality of
its services.  However, residential customer satisfaction increased,
then dropped, nationally and in 1994 returned to about the same level
as in 1991.  During this same period, the Service's use of
residential customer satisfaction data and related improvement
initiatives did not follow a sustained and well-coordinated national
strategy for improving customer service.  For example, the
improvements undertaken did not consistently focus performance
clusters on the most significant causes of customer
dissatisfaction--late and inconsistent mail delivery.  The Service
also did not include available measures of delivery performance, such
as on-time rates for overnight First-Class mail, in performance
incentive plans for executives and employees.  If the Service and its
employee organizations could agree to use such delivery measures,
along with measures now used, this might help to focus greater
corporatewide attention on improving those internal processes that
most affect the timeliness of mail delivery. 

After starting some national improvement initiatives, the Service did
not always make the best use of customer satisfaction data and other
available performance data to evaluate the initiatives.  Postal
headquarters units did not have a common approach for using the data
to measure and report on some national initiatives, such as efforts
to adjust post offices' hours to better recognize customer needs and
serve all post office customers in 5 minutes or less. 

At the conclusion of GAO's review, the Service was studying ways to
improve the use of all of its performance measures as part of a new
initiative called CustomerPerfect!sm.  Toward that end, GAO is making
several recommendations for improving the dissemination and the
potential use of customer satisfaction data by both Postal Service
management and Congress. 


   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4


      RESIDENTIAL CUSTOMER
      SATISFACTION DATA ARE SHARED
      WIDELY INTERNALLY BUT
      CONGRESS RECEIVES LITTLE OF
      THAT DATA
-------------------------------------------------------- Chapter 0:4.1

Postal leadership, particularly the Service's Consumer Advocate, who
reports to the Postmaster General, has made significant progress in
distributing and using residential customer survey results
internally.  This internal communication comes through oral
briefings, quarterly written reports, and automated information
systems.  The Consumer Advocate did various analyses of the data
every quarter for postal leadership.  Management at all levels has
access to up-to-date information on the satisfaction of these
customers through detailed hard-copy reports and automated corporate
information systems. 

The Postal Reorganization Act of 1970 calls for comprehensive
statements to be submitted by the Postal Service to Congress each
year on its plans, policies, and procedures for carrying out its
universal service mission.  The statements are to describe postal
operations generally and include, among other data, the speed and
reliability of service provided for the various classes of mail and
types of mail service.  However, the Postal Service provides
information in those reports on only 1 of 39 questions regarding
residential customer satisfaction nationally and excludes available
data on on-time delivery performance that are most critical to the
Service's sucess.  In 1993 and again in 1994, following its efforts
to downsize and reduce overhead costs, the Service reduced the amount
of customer satisfaction information and other performance data that
it provided in the required comprehensive statements to Congress. 


      BUSINESS CUSTOMER
      SATISFACTION DATA NOT
      DISTRIBUTED WITHIN POSTAL
      SERVICE OR TO CONGRESS
-------------------------------------------------------- Chapter 0:4.2

In 1993, the Service began gathering information on business customer
satisfaction under an $11.9 million contract.  However, the Service
has distributed no data on business customer satisfaction.  Under the
contract, the Service was to receive quarterly information beginning
in April 1994.  Postal Service management officials from the national
level down to performance cluster levels were to use the information
for allocating resources to maximize customer satisfaction and better
understand customer expectations and improve service. 

However, after being briefed by the contractor on the business
customer survey results, the Service became concerned that the data
could become available to its competitors.  The 1970 act allows the
Service to withhold data from the public that are of a commercial
nature.  Top Postal Service officials told the contractor not to
disseminate the information because of a concern that the information
might be publicly released. 

In a previous report,\1 GAO agreed with the Service that public
disclosure of detailed customer satisfaction data for specific
geographic areas or particular aspects of its services could harm its
interests.  That report also discussed the practice of some of the
Service's competitors of collecting and using customer satisfaction
data internally and releasing overall data to the public. 

GAO did not review the business customer satisfaction data compiled
by the contractor, but such data would seem useful to Postal Service
management for improving customer service and to Congress for its
oversight activities.  Both the Service and Congress have found
similar data on residential customer satisfaction useful.  Moreover,
business customers are important to the Service's business success
because they generate 88 percent of the Service's revenue, totaling
$49 billion in fiscal year 1994. 

Given its experience with the external distribution of overall
residential data, it appears that the Service could similarly share
some overall business customer satisfaction results with Congress. 
This could perhaps be done by presenting indicators of business
customer satisfaction nationally, for broad customer groupings and/or
for large geographic areas.  GAO believes that the Service should
consult with Congress to determine what data will best serve its
oversight needs. 


--------------------
\1 See U.  S.  Postal Service:  Tracking Customer Satisfaction in a
Competitive Environment (GAO/GGD-93-4, Nov.  12, 1992). 


      NUMEROUS SERVICE INITIATIVES
      UNDER WAY TO IMPROVE
      CUSTOMER SATISFACTION
-------------------------------------------------------- Chapter 0:4.3

Since 1990, the Postal Service has begun or expanded many different
efforts to improve customer service and reduce customer
dissatisfaction.  Its initiatives include programs to encourage,
train, and reward employees across the board to focus more on serving
customers.  For example, the Service trained over 100,000 retail
employees (postmasters, window clerks, and bulk mail acceptance
clerks) on courtesy and product knowledge.  It also set new policies,
such as the use of debit and credit cards, and new standards, such as
service in 5 minutes or less at post offices, to focus greater
corporatewide attention on customer service.  In line with national
customer service goals, the postal field offices that GAO visited
were all pursuing a broad array of local efforts to improve service. 

Even though the Service was pursuing numerous improvement efforts,
there still has been little overall change in the level of customer
satisfaction.  Nationally, 85 percent of all residential customers
rated the Service's performance excellent, very good, or good in
postal quarter 4, 1994 (May 28 to September 16).\2 This rating was
identical to the first quarterly rating in 1991, although the
national rating did increase but then decreased by 4 percentage
points during this period. 


--------------------
\2 The postal fiscal year starts and ends in September of each year
and includes a total of 13 four-week accounting periods.  Postal
quarters 1 through 3 include three accounting periods each, and
postal quarter 4 includes the last four accounting periods. 


      ADDITIONAL STEPS THAT MIGHT
      HELP FOCUS MORE ATTENTION ON
      IMPROVING CUSTOMER
      SATISFACTION
-------------------------------------------------------- Chapter 0:4.4

GAO identified opportunities for the Service to make better use of
customer satisfaction and other performance data for the purpose of
improving customer service.  Although many of the Service's
improvement initiatives began as early as 1990, it had not yet
developed and begun implementing at the performance cluster level a
strategy to integrate the initiatives and focus them on aspects of
the Service's operations that most relate to customer
dissatisfaction. 

Quarterly analyses done by the Consumer Advocate of customer
satisfaction data show that improving the reliability of mail
delivery offers the greatest opportunity to improve customer
satisfaction.  Managers and employees in mail processing plants, who
affect on-time delivery rates for most mail, were often less involved
than those at post offices in analyzing customer satisfaction data
and identifying actions that could lead to improved customer service. 
While managers and employees in these plants used other performance
measures, such as number of pieces left unprocessed each morning, the
use of those measures and customer satisfaction data could help to
emphasize the importance of timely mail processing to achieving
customer satisfaction. 

Further, performance pay incentives available to executives,
managers, and some postal employees did not incorporate available
measures of delivery service reliability, such as independent
measures of First-Class on-time delivery rates, for determining
annual changes in pay.  Continued use of these measures, along with
customer satisfaction and other performance measures, could help
encourage employees to make delivery service more reliable, and form
a basis for rewarding them.  Such action would require the
cooperation of unions and management associations within the Service. 

In September 1995, Service officials said that as part of a new
top-down, Service-wide effort called CustomerPerfect!, the Service
will introduce a new compensation plan for all postal executives for
fiscal year 1995 performance.  Incentive payments for executives are
to be determined on the basis of corporate financial performance,
timeliness of mail delivery, and employee satisfaction.  The
officials said that a similar incentive plan will be proposed in
future consultations with management associations and negotiations
with the unions. 

Finally, not all postal headquarters units overseeing service
improvement initiatives followed a systematic approach for (1)
monitoring field offices' progress in implementing certain national
initiatives and (2) sharing information on the best customer service
practices of post offices and processing plants.  Customer
satisfaction data and other available performance measurement data,
such as on-time delivery rates, could be used to a greater extent in
these efforts.  Such data could help in assessing the extent to which
changes in internal processes have improved customer satisfaction,
both overall and with particular aspects of service, within
performance clusters. 


      PERFORMANCE MEASURES UNDER
      STUDY
-------------------------------------------------------- Chapter 0:4.5

In November 1994, a high-level Service study group, including
officials who were briefed on the business customer survey results,
was studying how to use business customer satisfaction and other
performance data internally to improve customer satisfaction.  This
effort was part of the CustomerPerfect!  initiative to apply widely
accepted quality management criteria to enhance the Postal Service's
competitiveness.  No decision had been made as of September 1995 on
whether or when any of the business customer satisfaction data would
be disseminated within the Service and to Congress. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:5

GAO is recommending in chapters 2 and 3 that the Postmaster General
improve the dissemination and use of customer satisfaction and other
performance measurement data.  In particular, GAO recommends that the
Service use its ongoing study efforts to develop a plan, safeguards,
and timetable for distributing business customer satisfaction results
to all appropriate management levels of the Postal Service for use in
improving customer service.  As part of its study efforts, the
Service should also consult with appropriate congressional oversight
committees to determine what business and residential customer
satisfaction data and what other performance data the Service should
regularly provide to Congress for its use. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 0:6

The Postal Service provided written comments on a draft of this
report.  The comments are discussed at the end of chapters 2 and 3. 
The Service said that GAO's report presents a generally accurate
picture of what the Service was doing to measure customer
satisfaction and delivery performance at the time of GAO's review and
how the Service could better use the resulting data to improve
service quality. 

The Service believed that it had come a long way in the past 3 years
toward achieving its goal of developing a customer-driven,
customer-oriented, and customer-responsive organization.  The Service
said that its recent assessment of all functions and processes had
helped it to identify actions necessary to make that goal a reality. 
The Service believed that GAO's recommendations regarding information
sharing, employee performance incentives, systematic implementation
and monitoring of improvements, and sharing best practices will be
addressed in its newly established CustomerPerfect!  program.  For
example, the Service pointed out that as part of that program, a team
headed by the Consumer Advocate was studying the dissemination of
customer satisfaction results, both for business and residential
customers, and is to develop a strategy for making survey results
available to the public and Congress.  Another team was looking at
how to develop systems that will identify possible best practices and
validate their effectiveness by measuring their results. 

GAO believes that the Service's CustomerPerfect!  initiative appears
to be a reasonable approach to addressing its findings and
recommendations.  However GAO also notes that, at the time of it's
review, the Service had not obtained the involvement and commitment
of labor union leaders in the CustomerPerfect!  initiative.  GAO
believes that this involvement and commitment will be necessary to
most effectively implement the new initiative. 


INTRODUCTION
============================================================ Chapter 1

Most Americans use the U.S.  mail service.  Their opinions of that
service may depend on such factors as the timeliness of mail delivery
compared to their expectations, the time spent waiting in line for
window service, the availability of vending machines that work, and
the helpfulness of window clerks who are there to serve postal
customers. 

Concerned about untimely mail service at post offices, the Chairman
of the former Subcommittee on Information, Justice, Transportation,
and Agriculture, House Committee on Government Operations, asked us
to review selected aspects of the Postal Service's efforts to measure
and improve customer service.  Subsequently, we agreed to also report
the results of our review to the Chairman of the Subcommittee on the
Postal Service, House Committee on Government Reform and Oversight. 


   BACKGROUND
---------------------------------------------------------- Chapter 1:1

Under the Postal Reorganization Act of 1970, as amended, the Postal
Service is an independent establishment in the executive branch
operated as a basic and fundamental service provided to the people by
the government and accountable to Congress.  It is to provide
"prompt, reliable, and efficient services" to patrons in all areas
and render these services to all communities.  Over the years, the
Service has increasingly functioned as a businesslike entity
competing with electronic communication and private businesses to
provide communication and merchandise delivery services to
residential and business customers. 

Today, the Postal Service's customer base is diverse, and the quality
of mail service has many dimensions, such as whether the time to
deliver mail meets standards, access to service is convenient, and
service is timely and courteous at post offices.  Until recent years,
the Postal Service's measurement of service quality was internally
focused.  For example, it measured the time to process mail from
points within the postal system; it did not measure the time from
deposit of mail into the system to delivery of mail to customers. 

The Service's customer orientation continues to change. 
Increasingly, it is focusing on customer needs, expectations, and
perceptions.  Its two principal measures of service quality are the
Customer Satisfaction Index (CSI) and Business Customer Satisfaction
Index (BCSI), which measure how residential and business customers,
respectively, perceive the Postal Service's performance; and the
External First-Class Measurement System (EXFC), a quantitative
measure of total delivery performance.  Both measurements are done
independently of the Postal Service:  residential CSI by Opinion
Research Corporation; business CSI by Gallup Organization, Inc.; and
EXFC by Price Waterhouse and Co.  The Vice President/Consumer
Advocate, who reports directly to the Postmaster General, oversees
the CSI survey process and the EXFC end-to-end measurement system
and, until December 1994, was responsible for analyzing and
disseminating CSI and EXFC results.\1

The quality focus of the Postal Service leadership team is consistent
with current national objectives of making government more responsive
to the American public.  These objectives, outlined recently by the
National Performance Review (NPR) task force,\2 emphasize the need to
change the way government works by putting the customer first, giving
the customer a voice, and setting customer service standards.  The
Postal Service is following the NPR guidance by recognizing the need
to continuously improve customer service to remain competitive. 

Establishing and maintaining consistently high levels of delivery and
retail service are critical to the Service's success in an
increasingly competitive communications marketplace.  We previously
reported\3 that the Postal Service is losing profitable business to
the private sector, especially in the parcel post and overnight mail
markets.  Private carriers dominate the profitable
business-to-business segment because they offer cheaper and faster
service and have left the Postal Service with the more dispersed and
less profitable household market segment. 


--------------------
\1 In December 1994, the responsibility for analysis and internal
dissemination of CSI data was reassigned to the Vice President for
Work Force Planning and Service Management. 

\2 The NPR's report, From Red Tape to Results:  Creating a Government
That Works Better & Costs Less, was issued in September 1993. 

\3 U.S.  Postal Service:  Pricing Postal Services in a Competitive
Environment (GAO/GGD-92-49, March 25, 1992). 


      POSTAL SERVICE ORGANIZED FOR
      A CUSTOMER FOCUS
-------------------------------------------------------- Chapter 1:1.1

Soon after taking office in July 1992, the Postmaster General
outlined broad strategic goals that included improving service
quality and empowering employees to act responsively when customer
satisfaction is at stake.  In 1992, the Postmaster General downsized
the Postal Service but also reorganized it to focus greater attention
on serving customers.  The positions of vice president for customer
services and vice president for processing, both reporting to the
chief operating officer, were established at postal headquarters as
part of the reorganization.  The Postmaster General created 10 area
customer service offices, which oversee 85 customer service
districts, and 10 area processing offices, which oversee hundreds of
mail processing plants in the field. 

In 1994, to better coordinate customer service and mail processing
functions, the Postmaster General eliminated the two above-mentioned
vice president positions at headquarters.  He also combined the two
manager positions in each area into a single vice president position
responsible for both customer service and mail processing.  Below the
area level, district and plant managers continue to report separately
to the area vice president. 

Postmasters report to customer service district managers, who report
to an area vice president and oversee retail service operations of
about 40,000 post offices, stations, and branches nationwide.  Plant
managers report directly to an area vice president and oversee about
500 air, bulk, and general mail processing plants. 


      RESIDENTIAL CSI SURVEYS ARE
      DESIGNED FOR BOTH INTERNAL
      AND EXTERNAL NEEDS
-------------------------------------------------------- Chapter 1:1.2

The Postal Service serves 125 million households 6 days a week.  Its
residential customer surveys are done every postal quarter\4

to measure these customers' perceptions of virtually all postal
services.  Its surveys cover the Postal Service's 10 geographical
areas; 85 service areas, which include customer service districts and
processing plants (called "performance clusters"); and 170
metropolitan areas of the United States.\5 Under the $10.9 million
contract with Opinion Research Corporation, through December 1994,
the Service had received residential CSI results for 13 postal
quarters dating back to April 1991.\6 The results show the
perceptions of residential customers regarding the Service's overall
performance (question 1a of the CSI survey questionnaire) and other
aspects of U.S.  mail services (37 other questions) for the 3 months
preceding the survey.\7 (A copy of the CSI survey questionnaire is
included as app.  I.) Customers receiving the questionnaire are also
asked to provide written comments on (1) especially good experiences
with the Postal Service and (2) anything that the Service could do to
increase customer satisfaction.  We previously reported\8 that the
CSI surveys were designed to provide a statistically valid survey for
measuring customer satisfaction with the quality of postal services. 

The Service makes some residential CSI results publicly available
each quarter showing overall customer satisfaction nationally and for
the 170 metropolitan areas.  The results are to be used internally to
help track trends in customer satisfaction over time and by
organizational component.  The results also are to serve as a
diagnostic tool for improving the quality of both delivery and retail
services. 

In April 1993, the Postal Service awarded a 4-year $8.3 million
contract to The Gallup Organization, Inc., to develop and operate for
the Postal Service a Business Customer Satisfaction Index (BCSI)
measurement system.  Subsequent contract amendments increased the
estimated total cost to about $11.9 million, and the Service had
spent about $6.0 million under the contract through September 1995. 
The information from the system was to be used to measure the
satisfaction of these customers and determine the allocation of
resources needed to maximize customer satisfaction.  The system was
to produce valid and projectable data for each of 170 metropolitan
areas and provide for aggregating the data for performance clusters
and higher postal organizational levels. 


--------------------
\4 The postal fiscal year starts and ends in September of each year
and includes a total of 13 four-week accounting periods.  Postal
quarters 1 through 3 include three accounting periods each, and
postal quarter 4 includes the last four accounting periods. 

\5 Subsequent to our review, for its reporting purposes, the Service
consolidated two metropolitan areas, Anchorage and Juneau, Alaska,
which changed the total from 170 to 169. 

\6 The Postal Service also did CSI surveys for the first 3 quarters
of fiscal year 1991, but the surveys did not cover some of the 170
metropolitan areas that were covered beginning with postal quarter 4,
fiscal year 1991.  Before the CSI surveys, the Service used other
measures of customer satisfaction.  One of these was a semiannual
National Tracking Study that began in 1973 in which the general
public rated its satisfaction with the Postal Service.  In April
1990, the overall favorable rating was 83 percent, which, the Service
reported to Congress that year, "approached its all-time high."

\7 CSI surveys are generally mailed very early in a postal quarter. 
Many of the items on the questionnaire ask customers to respond on
the basis of their experience "during the past three months." All of
the quarterly CSI data in this report are for the postal quarter for
which customers provided their perceptions, not the quarter in which
the surveys were mailed and returned. 

\8 U.S.  Postal Service:  Tracking Customer Satisfaction in a
Competitive Environment (GAO/GGD-93-4, Nov.  12, 1992). 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:2

The Chairman of the former Subcommittee on Information, Justice,
Transportation, and Agriculture, House Committee on Government
Operations, requested that we review selected aspects of the Postal
Service's efforts to measure, report, and improve customer
satisfaction.  Subsequently, the Chairman of the Subcommittee on
Postal Service, House Committee on Government Reform and Oversight,
requested that we report the results of our review to that
Subcommittee. 

Our objectives were to determine (1) to what extent the Postal
Service disseminates residential and business CSI data internally and
to Congress and (2) whether opportunities exist for the Postal
Service to improve the dissemination of CSI data and their potential
use by Congress and the Postal Service.  We were also to determine
(3) the steps that the Postal Service is taking to improve customer
satisfaction using CSI and other data and (4) any additional steps
the Service could take to improve customer service and thereby
improve customer satisfaction. 

Because the Postal Service had not made public any BCSI data at the
time of our review, our work on the dissemination and use of customer
satisfaction data was limited to residential CSI data.  We reviewed
the Gallup contract, analyzed Postal Service data on the relative
importance of residential and business mail to the Service's overall
mail volumes, and obtained explanations from Postal Service officials
of the status of the Gallup contract and plans for dissemination of
BCSI data. 

As part of our work on residential CSI data collection,
dissemination, and use, we interviewed Postal Service headquarters
officials, including the Chief Operating Officer and the Vice
Presidents for Customer Services, Consumer Affairs, and Quality.  We
interviewed various other headquarters officials responsible for
customer retail service to find out how residential CSI data were
used and what improvement initiatives were under way.  We reviewed
various materials and documents, such as reports, video tapes, and
briefing documents, used by postal headquarters and selected field
offices to disseminate CSI data.  We analyzed annual reports sent by
the Postal Service to Congress during fiscal years 1991 through 1994
as part of our efforts to determine any opportunities for the Service
to improve the sharing of information on customer satisfaction and
its performance with Congress. 

Along with interviews with numerous headquarters and field postal
officials, we reviewed CSI-related reports prepared by Opinion
Research Corporation and the Postal Inspection Service to identify
opportunities to improve the dissemination and potential use of CSI
data.  We used the results of all of these tasks to assess the
planning and monitoring of initiatives undertaken to improve customer
satisfaction by improving customer service. 

To determine the extent of improvement in customer satisfaction with
postal services, we obtained CSI metropolitan area data\9 on question
1a responses relating to customers' perceptions of satisfaction with
the "overall performance" of the Postal Service and 22 other
questions on window, telephone, and related retail services.  We also
analyzed EXFC data to determine changes in on-time delivery rates
since the measurements began in 1990 and compared EXFC data with CSI
data for the nation and selected metropolitan areas. 

We estimated sampling errors for the CSI results in each of 170
metropolitan areas using CSI data provided by the Service for the
first quarter of fiscal year 1992 through the third quarter of fiscal
year 1994.  Estimates of sampling errors for each area are based on
simple random sampling assumptions.  Sampling errors are not reported
for specific metropolitan areas because the Postal Service did not
provide us with the names of specific metropolitan areas associated
with the data.  We also analyzed national and metropolitan area CSI
results from the first quarter of fiscal year 1991 through the fourth
quarter of fiscal year 1994.  We could not calculate sampling errors
for national CSI results using the data provided. 

The Postal Service uses 95-percent confidence intervals as indicators
of sampling errors for percentages.  This means that the chances are
about 95 out of 100 that the actual percentage falls within the
confidence interval.  For example, if 83 percent are reported to be
satisfied with Postal Service performance and the sampling error is
reported to be ï¿½ 3 percentage points, the chances are about 95 out of
100 that the actual percentage satisfied is between 80 and 86
percent. 

To identify steps that the Service is taking to improve customer
satisfaction using CSI and other data, we held numerous interviews at
postal headquarters and selected postal field offices and, as
appropriate, obtained supporting documentation.  We visited six
customer service districts--three having among the highest CSI
ratings for retail services in the nation (Billings, MT; Central
Plains in Omaha, NB; and Springfield, MA) and three having some of
the lowest CSI scores for retail services (Chicago, IL; New York, NY;
and San Francisco, CA).  Our purpose in selecting a mix of
high-scoring and low-scoring districts was to identify innovative
service improvement initiatives in some districts with different
levels of customer satisfaction. 

We interviewed the six area managers for customer services with
responsibility for the six districts we visited.  We held discussions
with the district manager and his/her key assistants, the consumer
affairs manager, representatives of employee groups, retail
specialists, and selected postmasters and/or station managers.  We
toured several post offices or stations in each district.  Appendix
II presents background data on the six districts. 

Our review followed generally accepted government auditing standards. 
Our visits to postal field offices were made between November 1993
and May 1994.  For a significant period of time during our review, a
portion of our work was delayed because we did not have access to CSI
data needed to analyze customer satisfaction with retail services. 
We requested the data needed for this work in January 1993.  In
February 1994, the Postal Service provided the data we requested, and
we were then able to complete our CSI data analysis.  Our analysis of
CSI data was done between April 1994 and October 1995. 

We received written comments on a draft of this report from the U.S. 
Postal Service.  Summaries of these comments and our evaluation are
included at the end of chapters 2 and 3.  The comments are reprinted
in appendix VI. 


--------------------
\9 The Service's CSI sampling plan provides for random selection of
households from each of 170 metropolitan areas each postal quarter. 
On average, about 1,200 households in each metropolitan area return
questionnaires that are used for reporting customer satisfaction
nationally and for 10 postal area offices, 85 performance clusters,
and numerous 3-digit ZIP Code areas.  Our earlier report on gathering
CSI data, referenced in footnote 8, discusses the possible bias in
interpreting CSI results because of low response rates and the
Service's efforts to monitor the situation. 


OPPORTUNITIES EXIST TO IMPROVE THE
DISSEMINATION AND POTENTIAL USE OF
CUSTOMER SATISFACTION DATA
============================================================ Chapter 2

The Postal Service makes extensive internal dissemination of
residential CSI data to track customer satisfaction and identify
opportunities to improve customer service.  However, it has not
disseminated much of that data to Congress and recently further
limited the data provided in required reports to Congress.  Moreover,
the Service has gathered business CSI information but has not
disseminated it internally or to Congress.  Along with improving CSI
data dissemination, the Postal Service can potentially improve use of
the residential data by giving field offices more guidance on
analyzing certain CSI results. 


   RESIDENTIAL CSI RESULTS ARE
   WIDELY DISSEMINATED WITHIN THE
   POSTAL SERVICE
---------------------------------------------------------- Chapter 2:1

Postal Service officials believe that the use of residential CSI
results can help improve organizational and employee commitment to
customer satisfaction.  Accordingly, the Consumer Advocate and other
officials have taken numerous steps to make Postal Service leadership
and employees aware of and help them use CSI results.  Soon after
residential CSI results became available each quarter, the Consumer
Advocate provided the results to the Postmaster General and other top
postal leaders.  The results were disseminated widely within the
organization in several ways: 

  -- The Postmaster General highlighted CSI results in his quarterly
     report to the Board of Governors. 

  -- The Consumer Advocate provided more detailed briefings for the
     Board of Governors on the survey results each quarter,
     highlighting customers' ratings of the Service's overall
     performance and identifying the highest and lowest ranked
     customer service districts.  The Consumer Advocate also visited
     postal facilities in metropolitan areas having the highest
     rating for the quarter to commend local management and
     employees. 

  -- The Service's contractor, Opinion Research Corporation, provided
     quarterly written reports detailing CSI results for use by
     postal headquarters and each subordinate management level in
     Washington, DC, and field locations. 

  -- The Postal Service made CSI results available electronically to
     executives, managers, and employees through automated
     information systems. 

To further promote the use of CSI results, in November 1992, the
Consumer Advocate established an Independent Service Analysis Group\1
to assist offices and individuals throughout the Service in using CSI
and other customer service data, such as EXFC and customer complaint
data.  The group made various analyses and issued reports of CSI
results periodically and on demand for postal leadership, including
the Board of Governors, and postal managers at all levels.  Each
quarter the group identified the top 10 and bottom 10 metropolitan
areas of the total 170 metropolitan areas.  The group also made
comparisons each quarter to show whether and to what extent each
performance cluster's CSI ratings differed from (1) the current
median rating for all clusters and (2) the cluster's rating for the
same quarter of the previous year.  The Service made available to all
management levels, through an automated information system, the
results of these comparisons for those performance clusters and CSI
questions having significant differences from the prior year. 

Along with data analysis, the group provided instructions to field
offices on how to use CSI results for changing internal processes
that caused customer dissatisfaction.  The manager and other members
of the group had visited all 10 area offices and numerous district
offices to assist with CSI data analysis.  The group also developed
and furnished video tapes on how to analyze and use CSI data. 


--------------------
\1 In December 1994, the group and its function were transferred to
the Vice President for Work Force Planning and Service Management. 


      FIELD OFFICES FURTHER
      DISSEMINATED CSI RESULTS
-------------------------------------------------------- Chapter 2:1.1

At the time of our review, Postal Service headquarters had not
prescribed specific procedures and methods for area, district, and
processing plants to use in disseminating and using CSI results.  The
three area offices and six districts that we visited used a variety
of means to provide CSI results to managers, supervisors, and
employees.  For example, the Central Plains District in Omaha, NB,
published Newsbreak, a monthly information newsletter for its
employees that periodically included CSI information.  The Pacific
Area Office in San Francisco, CA, had prepared a 12-minute videotape
to be shown to employees in which area managers provided an overview
of the CSI process and employees then spoke about their roles in
improving customer service. 

District managers said that they discussed CSI results in regular
meetings with postmasters and employees.  Local managers said that
they found narrative comments included in CSI reports to be
especially useful because the managers could identify customer
concerns about service at specific post offices.  For example,
customer service managers in the New York District said that every
quarter they analyze hundreds of narrative comments made by customers
to better understand customers' perceptions of service quality in
specific locations. 


   LITTLE RESIDENTIAL CUSTOMER
   SATISFACTION DATA PROVIDED TO
   CONGRESS
---------------------------------------------------------- Chapter 2:2

Although the Postal Service has generated valuable information from
its residential customer surveys since 1991, it has provided
relatively little of the information to Congress.  The Postal Service
also shares very little information on residential customer
satisfaction with the public.  In recent years, the Service has
reduced the amount of residential customer satisfaction data and
other performance data provided in required annual comprehensive
statements to Congress. 

The Service publicly discloses the responses to only 1 CSI question
on the Service's overall performance for the nation and 170
metropolitan areas each quarter.  The Service included data on only
this one CSI question in comprehensive statements to Congress that
are required annually by the 1970 act (39 U.S.C.  2401). 

Provisions of the act calling for comprehensive statements specify
several categories of data to be included in each statement.  The
statements are to cover the Service's plans, policies, and procedures
for carrying out its universal mail service mission, which is stated
in section 101 of the act.  The statements are to also describe
postal operations generally and include data on the speed and
reliability of service provided for the various classes of mail and
types of mail service, mail volume, productivity, trends in postal
operations, and analyses of the impact of internal and external
factors on the Postal Service.  The act also says that the Senate and
House postal oversight Committees of Congress are to hold hearings on
the Postal Service in March each year. 

As a stakeholder in the delivery of U.S.  mail, Congress has not only
described in the 1970 act certain information it needs from the
Service but also has often expressed interest in particular aspects
of the Service's performance and customer satisfaction.  In 1994,
this interest was manifested in congressional hearings and public
statements of several Members of Congress regarding the quality of
service in the Washington, DC, and Chicago, IL, areas.  Typically,
Members of Congress have responded after news accounts and complaints
from the public regarding the quality of mail service in particular
areas of the country.  The Postal Service reacted to concerns about
the quality of delivery service in testimony before Congress several
times in 1994.  More recently, the House oversight committee has held
numerous general oversight hearings with a view toward determining
the need for any changes in the 1970 act. 


      POSTAL SERVICE REDUCED
      AMOUNT OF CSI AND OTHER
      PERFORMANCE DATA IN
      COMPREHENSIVE STATEMENTS
-------------------------------------------------------- Chapter 2:2.1

The Postal Service has submitted the required comprehensive
statements, and the oversight Committees have held hearings on the
Postal Service's operations and services.  However, the usefulness of
the comprehensive statements has been limited by the scant CSI and
other performance data included in them, particularly in the
statements for fiscal years 1993 and 1994. 

Our review of the last four statements (fiscal years 1991 through
1994) showed that the Postal Service has reduced the amount of
information on customer satisfaction and delivery performance
provided to Congress.  The 1992 statement tabulated on-time delivery
rates from EXFC for overnight, 2-day, and 3-day delivery service for
each quarter and the year.  The 1992 statement also included CSI
results for each quarter, with the results broken out into several
categories of customer responses, i.e., excellent only; excellent and
very good combined; good only; excellent, very good, and good
combined; fair only; poor only; and fair and poor combined. 

Following the Service's efforts to downsize and reduce overhead
costs, the 1993 statement to Congress provided less CSI information
than the 1992 statement.  The 1993 statement showed only one rating
for the year, which included the sum of all excellent, very good, and
good responses for only the fourth quarter of 1993.  For comparison
purposes, the sum of these same responses was included for the fourth
quarter of 1992.  The 1994 statement had even less CSI information
than the 1993 report.  The 1994 statement showed one rating
(excellent, very good, and good combined), and it was for the fourth
quarter of fiscal year 1994 only.  No comparison was presented of the
1994 rating with the same quarter of 1993. 

The Postal Service provided more CSI information to the general
public than was provided in its required comprehensive statements to
Congress.  In quarterly publications available to the public, the
Service included CSI ratings for 170 metropolitan areas.  In
addition, the publications also provided the EXFC ratings for 96 of
these same areas.  Further, the Service recently added new
measurements of its on-time delivery performance for the mailings of
those publishers and mailers participating in the external
second-class and third-class mail measurement systems.  The Service
is also participating with some foreign postal administrations in the
development of processes for measuring international air letters.  No
data were provided in any of the four comprehensive statements that
we reviewed relating to these new measurements. 


      ANALYSIS OF CSI RESULTS AND
      OTHER PERFORMANCE DATA COULD
      BE MORE INFORMATIVE
-------------------------------------------------------- Chapter 2:2.2

The Service could use data that it already routinely releases to the
public and that it previously provided to Congress for more
informative analysis and presentations in required comprehensive
statements to Congress.  Use of these CSI results and other
performance indicators could provide a more complete picture of
customer satisfaction and the Service's performance.  No additional
data-gathering would be necessary.  CSI and EXFC results are
available to the public in quarterly publications prepared by the
Consumer Advocate, but the results are not compared, analyzed, and
summarized for potential use by Congress.  The following illustrates
some ways in which the Postal Service might present additional CSI
and EXFC results to Congress.  Although the Service compiles data
similar to EXFC for other mail classes, we did not include data on
these classes because the Service does not release any of that data
to the public. 


         CSI RESULTS COULD BE
         PRESENTED FOR PAST
         PERIODS AND BY GEOGRAPHIC
         AREAS
------------------------------------------------------ Chapter 2:2.2.1

Congress could use comparisons of CSI results for several years to
review the progress the Postal Service has made in improving
perceptions of its overall performance.  The Postal Service's annual
report for 1994 did not present comparative CSI results for the
current year and preceding year.  In future annual reports and other
communications with Congress, the Service could use available CSI
data that are currently or were previously made public to compare
customer satisfaction, by postal quarter, for the current and
previous years.  (See fig.  2.1.)

   Figure 2.1:  Comparison of
   National CSI Ratings, by Postal
   Quarter, Fiscal Years 1993 and
   1994

   (See figure in printed
   edition.)

Source:  CSI Surveys, U.S.  Postal Service. 

The Postal Service could also use tables or graphics to show how the
results differ among the 10 postal area offices and 85 performance
clusters.  CSI ratings differ significantly among geographic areas of
the country, and with better disclosure of the ratings, Congress
could compare and contrast customer satisfaction levels and changes
not only nationally but also for various regions and cities.  For
example, the Postal Service could show how the results compare among
the 10 postal areas for selected periods.  (See fig.  2.2.)

   Figure 2.2:  Overall
   Satisfaction for 10 Postal Area
   Offices for Postal Quarter 4 of
   Fiscal Years 1993 and 1994

   (See figure in printed
   edition.)

Note:  CSI percentages include excellent, very good, and good
ratings. 

Source:  CSI Surveys, U.S.  Postal Service. 

CSI results could also be presented for the Service's performance
clusters.  For its internal use, the Postal Service arrays CSI
results by cluster and compares current and preceding year results. 
Thus, the Service would need little additional effort to include such
information in required annual reports to Congress.  Although the
Service does not currently present CSI data to the public in this
manner, it does publish CSI ratings every quarter on smaller
geographic areas--the 170 metropolitan areas.  Presenting the data
for larger geographic areas would not appear to pose any greater
threat to the Service's competitive interests than disclosing the
data by metropolitan areas, as is done now. 


         DIFFERENCES IN CSI AND
         EXFC RESULTS COULD BE
         PRESENTED AND EXPLAINED
------------------------------------------------------ Chapter 2:2.2.2

CSI and EXFC are the Service's two most widely publicized externally
developed performance measures.  Although the two systems are very
different and so are the results, the Service presented quarterly
ratings from the two systems, both for metropolitan areas and
nationally, side by side in its publicly disseminated documents.  The
Service could present CSI and EXFC data in a way that helps ensure
that the extent of and reasons for differences in customer perception
of the Postal Service's performance and measurement of delivery
performance are understood. 

The Service has tended to focus much of its attention on publicizing
customers' perceptions of the Service's overall performance and
improving these perceptions.  While the publicized ratings disclose
perceptions of overall performance, various data compiled by the
Service show that customers are most concerned about the length of
time that the Service takes to deliver mail and the consistency of
mail delivery service.  However, the Service recognizes that a number
of factors, not necessarily related to mail delivery, influence
customer perception of the Service's overall performance, as measured
by CSI surveys. 

To illustrate how performance perceptions can differ from delivery
measurements, in postal quarter 4, 1994, EXFC scores for 28 of 93
metropolitan areas varied by more than 5 percentage points from CSI
scores for the same areas.\2 For 16 of the 28 areas, the CSI ratings
were higher than the EXFC ratings.  For 8 of these same 28
metropolitan areas, the difference between the EXFC and CSI ratings
was 10 or more percentage points.  For five of these eight, the CSI
ratings were higher than the EXFC ratings.  (See fig.  2.3)

   Figure 2.3:  Comparison of CSI
   and EXFC Ratings for 8
   Metropolitan Areas With
   Differences of 10 or More
   Percentage Points in Postal
   Quarter 4, Fiscal Year 1994

   (See figure in printed
   edition.)

Note:  CSI percentages include excellent, very good, and good
ratings. 

Source:  CSI Surveys and EXFC System, U.S.  Postal Service. 

For the remaining three metropolitan areas (Chicago, IL; Queens, NY;
and Washington, DC), the CSI ratings were lower than the EXFC ratings
in postal quarter 4, 1994.  Of the 170 metropolitan areas for which
CSI results are reported, these 3 areas were the 3 lowest ranked in
quarter 4.  In all three of these metropolitan areas, the CSI ratings
had dropped below the EXFC rating during 1994.  The EXFC results
improved in all three areas during the year, but the CSI ratings for
all three were still well below their EXFC ratings at year's end. 
(See figs.  2.4-2.6.)

   Figure 2.4:  District of
   Columbia CSI and EXFC Ratings
   by Postal Quarter, Fiscal Year
   1994

   (See figure in printed
   edition.)

Note:  CSI percentages include excellent, very good, and good
ratings. 

Source:  CSI Surveys, U.S.  Postal Service. 

   Figure 2.5:  Chicago, IL, CSI
   and EXFC Ratings by Postal
   Quarter, Fiscal Year 1994

   (See figure in printed
   edition.)

Note:  CSI percentages include excellent, very good, and good
ratings. 

Source:  CSI Surveys, U.S.  Postal Service. 

   Figure 2.6:  Queens, NY, CSI
   and EXFC Ratings by Postal
   Quarter, Fiscal Year 1994

   (See figure in printed
   edition.)

Note:  CSI percentages include excellent, very good, and good
ratings. 

Source:  CSI Surveys, U.S.  Postal Service. 

As indicated above, for some areas, customer perception of overall
performance remained relatively low after overnight First-Class
delivery performance improved.  Because of such differences, it is
important that the results of the CSI and EXFC systems be presented
in a way that makes clear that they represent two very different
measures of the Service's performance. 

A Postal Service manager responsible for CSI data analysis said that
the Service does not expect a "match" between CSI and EXFC results,
either overall or by specific CSI question or service attribute.  He
said there is a tenuous relationship between internally driven
commitments, e.g., overnight delivery service, and customer
expectations.  He said that responses to CSI question 1a are affected
by many factors, such as the Service's announcements of postage rate
increases and adverse publicity in the news media, and that on-time
delivery explains about one-half of the question 1a results.  After
reviewing a draft of this report, the Vice President/Consumer
Advocate agreed with the manager's comments summarized above.  She
said, however, that the impact of adverse publicity on CSI ratings is
short-lived and does not affect the ratings in every metropolitan
area across the nation. 

We believe that the above comments by the manager and Vice President
are all good reasons why CSI and other performance data need to be
analyzed and presented to Congress in a way that provides as complete
and accurate a picture as possible of both the Service's delivery
performance and customer perceptions of its performance. 


--------------------
\2 The Service considers any difference of more than 3 percent to be
statistically significant. 


         CUSTOMER RESPONSES COULD
         BE BROKEN OUT TO BETTER
         DISCLOSE LEVELS OF
         SATISFACTION
------------------------------------------------------ Chapter 2:2.2.3

In presenting CSI results to Congress, the Postal Service could break
out the results to show more clearly how satisfied, in terms of
specific response categories, customers are with the Service's
overall performance.  As stated previously, in some earlier reports
to Congress such breakouts were provided. 

Customers can rate the Service's performance as excellent, very good,
good, fair, and poor.  The Postal Service disclosed only what it
termed "favorable" responses when presenting CSI information in
quarterly pamphlets to the public.  These responses were the sum of
excellent, very good, and good responses for each of 170 metropolitan
areas and the nation.  If the Service disclosed the percentage of
customers giving the higher ratings of excellent and very good
combined, as it does for its internal reports, Congress would have a
more precise picture of how customers' perceptions have changed over
time. 

For most of its internal purposes, including calculation of
performance incentive payments for executives and employees
(discussed in ch.  3), the Service uses excellent and very good
ratings only.  Most of its management reports use these two ratings
alone or in combination with the overall favorable rating, which
includes not only excellent and very good responses but also good
responses.  Disclosing excellent and very good responses is important
because, as figure 2.7 shows, good responses alone accounted for
almost one-third of all responses. 

   Figure 2.7:  Good Responses
   Accounted for Almost One-Third
   of Customers' Responses in
   Fiscal Years 1992, 1993, and
   1994

   (See figure in printed
   edition.)

Source:  CSI Surveys, U.S.  Postal Service. 

Combining good responses with excellent and very good responses and
reporting only the totals can mask shifts in customer satisfaction,
and the changes can sometimes be statistically significant.  This
masking occurs when customers either increase their ratings from good
to the higher ratings or drop ratings from excellent and very good to
good.  For example, for the San Francisco, CA, metropolitan area, the
favorable rating increased by 4 percentage points, from 82 to 86,
which the Service considers to be statistically significant, between
quarter 4, 1993, and quarter 4, 1994.  However, excellent and very
good responses did not have a significant change, decreasing by 1
percentage point.  Conversely, in 14 metropolitan areas, the
percentage of excellent and very good responses together increased
even though the overall favorable rating went down between these same
2 quarters. 

Service officials include good responses in the publicized CSI
ratings and disclose that excellent, very good, and good responses
are combined.  However, the sum of excellent and very good ratings
alone, or together with the good ratings, nationally and for each of
the 10 postal areas and 170 metropolitan areas, would give Congress a
more complete picture of the status of and changes in customer
satisfaction.  In addition, such presentation would be more
consistent with the Postal Service's internal reporting.  This
further breakout of customer responses would not appear to jeopardize
the Service's commercial interests because the favorable ratings are
already available to the public. 


   IMPROVEMENT POTENTIAL EXISTS
   FOR USE OF RESIDENTIAL SURVEY
   RESULTS
---------------------------------------------------------- Chapter 2:3

Along with providing more comprehensive CSI information to Congress,
the Postal Service could potentially improve the usefulness
internally of residential customer survey results.  CSI results for
some questions have lower levels of precision.  While the Service has
taken steps to inform CSI users of the level of precision, written
reports distributed by the CSI contractor do not fully disclose the
level of precision. 


      INTERNAL USERS OF CSI
      REPORTS MAY NOT FULLY
      UNDERSTAND THE LIMITED
      PRECISION OF SOME RESULTS
-------------------------------------------------------- Chapter 2:3.1

Quarterly reports distributed by the CSI contractor to the Postal
Service contain extensive CSI data, including various satisfaction
percentages for 38 questions detailed to 3-digit ZIP code areas and
metropolitan areas.  The reports indicate which of these percentages
are significantly higher or lower than the national results. 
However, the usefulness of some of the percentages is limited by the
lower levels of precision.  While the Service had provided some
guidance to users on this data limitation, users of the
contractor-generated reports may not be sufficiently aware of how to
use those percentages having lower levels of precision.  The reports
give little guidance on how to interpret CSI data that are not as
precise as some other CSI data in the same reports. 

The Postal Service requires that the contractor survey enough
customers in each metropolitan area each quarter to provide a margin
of "sampling error"\3 associated with the responses for each CSI
question that is to be no greater than ï¿½ 3 percentage points for
question 1a on the CSI survey questionnaire.  Our review of response
rates for the 11 postal quarters from postal quarter 1, 1992, through
postal quarter 3, 1994, showed that the Postal Service obtained the
number of responses necessary to provide this required precision each
quarter. 

However, CSI results for some questions sometimes have sampling
errors that are much greater than ï¿½ 3 percent.  This occurs because
customers who have not used a particular service are instructed not
to answer questions about that service.  Because of this, the number
of responses for such questions, 22 of 38 in total, can be much
lower, and the sampling error much higher, than for question 1a. 

For example, customers who do not have any of their household's mail
delivered to a post office box are instructed not to answer the two
questions on this service.  In one metropolitan area, satisfaction
with delivery of mail to the correct post office box was 69 percent
in 1 quarter and 79 percent in another metropolitan area for the same
quarter.  However, both ratings were based on a small number of
responses:  29 responses in 1 metropolitan area and 36 responses in
the other.  The small number of responses results in a large margin
of sampling error.  (Details on sampling errors for metropolitan
areas are included in app.  III.)

Postal managers and employees are expected to use all CSI reports for
tracking progress in improving customer service and analyzing
processes at post offices and processing plants that affect customer
satisfaction.  However, high rates of sampling error for some
questions can result in inappropriate inferences if users of CSI
results compare one metropolitan area with another.  To illustrate
how this can happen, we will use the above example involving post
office box services.  After sampling errors are considered in this
case, the rating is between 53 and 85 (69 percent,
ï¿½ 16 percentage points) for one area and between 63 and 95 percent
(79 percent ï¿½ 16 percentage points) for the other area.  Thus, an
inference that the 79 rating indicates higher satisfaction than the
69 rating may be inappropriate because the difference could be due to
sampling error. 

Consumer Advocate officials we contacted were aware that CSI results
for some questions do not have the same degree of precision as the
overall ratings that are published quarterly using question 1a
responses.  They said that they inform field personnel of this
imprecision during all briefings and that the Corporate Information
System shows whether changes in responses to each question are
statistically significant.  They believe that even the less precise
results for some CSI questions can still be useful to managers,
particularly when combined with other data and when the results are
compared for several postal quarters or several years. 

We agree that the CSI results can be useful but also believe that
users of written CSI reports could be provided with additional
information on how sampling errors limit the precision of some CSI
data.  Because such errors can vary depending on the question, users
might benefit from additional information in the reports on the
sampling error for each CSI question.  This might help to ensure more
informed comparisons between metropolitan areas and over time. 


--------------------
\3 Sampling error is a measure of survey precision.  The smaller the
sampling error associated with a survey percentage, the more precise
it is.  For additional details on the effects of sampling error, see
Objectives, Scope, and Methodology. 


   BUSINESS CUSTOMER SATISFACTION
   DATA NOT DISSEMINATED IN POSTAL
   SERVICE OR TO CONGRESS
---------------------------------------------------------- Chapter 2:4

Although the Service disseminated CSI results for residential
customers, these customers represent a small portion of the Service's
mail volume.  Under a contract with the Service, The Gallup
Organization, Inc., gathers data on the satisfaction of business
customers, who account for the vast majority of the mail.  However,
the Service has not disseminated the results within the Service or
shared any results with Congress.  The Service is concerned that the
data might be made public if brought into the organization, thereby
jeopardizing its competitive interests. 


      BUSINESS CUSTOMERS GENERATE
      ALMOST 90 PERCENT OF MAIL
      VOLUME
-------------------------------------------------------- Chapter 2:4.1

It is important that the Service regularly obtain, analyze, and use
BCSI results, which are currently being generated for each postal
quarter, because business customers account for most of the Service's
mail volume.  Further, Postal Service studies show that these
customers are more likely than residential customers to switch to
another supplier of mail services. 

Most mail is a consequence of business transactions, including
billings, payments, advertising, and other economically motivated
activities.  Studies by the Postal Service show that 59 percent of
the total mail stream originates outside households and is sent to
households.  An additional 30 percent originates outside households
and is sent to nonhouseholds.  Overall, the flow of mail from
nonhousehold customers accounts for almost 90 percent of the total
mail.  (See fig.  2.8.)

   Figure 2.8:  Business-Generated
   Mail Was Almost 90 Percent of
   the Total Mail Stream in May
   1994

   (See figure in printed
   edition.)

Source:  U.S.  Postal Service Household Diary Study, May 1994. 

Similarly, almost 90 percent of the Service's revenue, totaling $58
billion in fiscal year 1994, is generated by business mailers.  The
level of satisfaction of these mailers and their continued use of the
U.S.  mail service are critical to the Postal Service's financial
viability.  Residential customers have limited alternatives for
letter mail service because of the Private Express Statutes\4
requiring the delivery of nonurgent letters by the Postal Service. 
But business customers often can and do use other private carriers
because of urgent delivery requirements, which are exempted from the
Private Express Statutes, and because their mailings are not what the
Service defines as letters for Private Express Statutes purposes. 


--------------------
\4 These statutes restrict the delivery of most letters to the U.S. 
Postal Service unless certain conditions that are specified in law
and postal regulations are met. 


      BCSI DATA ARE GATHERED BUT
      NOT REPORTED TO THE SERVICE
-------------------------------------------------------- Chapter 2:4.2

The Service began planning for quarterly surveys of business customer
satisfaction in 1991 and awarded a contract for the surveys to Gallup
in April 1993.  Under the contract, the contractor is required to
provide information for use by Postal Service management from the
national level down to the performance cluster level.  Management was
to use the information to determine the allocation of resources
needed to maximize customer satisfaction and analyze how to better
understand customer expectations and improve service.  During the
first year of the contract, Gallup was to conduct research and do a
pilot test of the measurement system.  The results of the first BCSI
survey were available to the Postal Service in April 1994.  Through
June 1995, Gallup said it had completed five quarterly surveys, but
the Service had not obtained and disseminated data from any of the
surveys for use in improving customer satisfaction. 

According to the former Consumer Advocate, who left the Service in
December 1994, some top Postal Service officials were briefed at
least once by the contractor on the BCSI results in 1994. 
Subsequently, the contractor was directed not to provide the
quarterly BCSI results to the Postal Service.  She said that the
Gallup surveys had produced "rich" data on business customer
satisfaction, which she believed postal management could use to
improve customer service.  Under the contract, Gallup was to provide
quarterly BCSI reports to all 170 performance clusters, 10 area
offices, and the Service headquarters.  Postal Service officials
confirmed that these required reports were not being submitted by
Gallup. 


      THE SERVICE IS CONCERNED
      ABOUT HARMING ITS
      COMPETITIVE INTERESTS
-------------------------------------------------------- Chapter 2:4.3

Service officials believe that the indiscriminate sharing of customer
satisfaction information with Congress and the public can be
self-defeating.  We agree that the Service's commercial interests
could be harmed by indiscriminate sharing where there is competition
for its services.  While most of the Service's mail volume is
protected by the Private Express Statutes, private companies compete
with the Service to provide certain mail services, particularly
expedited and parcel delivery, to residential and business customers. 
Competitors might use customer satisfaction and other performance
data, which the Service had gathered to improve its service and
become more competitive, to gain a competitive advantage over the
Service.  The 1970 act allows the Service to withhold from the public
data that are of a commercial nature. 

In particular, competitors might benefit from detailed CSI results
showing specific aspects of service and particular geographic areas
of the country where the Postal Service is not meeting customer
expectations.  Competitors could target their market development
efforts to these areas.  We previously reported\5 that the Service's
decision to publicly report only overall residential CSI ratings, but
not ratings of specific services, is permitted under the Postal
Reorganization Act of 1970.  Although the Postal Service is covered
by the Freedom of Information Act (FOIA) (5 U.S.C.  552), the 1970
act does not require it to disclose information of a commercial
nature, including trade secrets, that under good business practice
would not be disclosed publicly. 

In our earlier report, we also discussed the practices followed by
some of the Service's competitors in measuring and reporting customer
satisfaction.  The four competitors we contacted (Federal Express,
United Parcel Service, Associated Mail and Parcel Centers, and
Tribune Alternative Delivery) used independent contractors to assess
customer satisfaction.  Their goal was to achieve 100 percent
customer satisfaction for the specialized services they offered.  In
the highly competitive overnight and parcel business, only a customer
rating of "completely satisfied" (very good and excellent) was
acceptable to private carriers.  These companies did not release
detailed information on their customer satisfaction surveys because
they believed the information would be used to the advantage of their
competitors.  One of the companies, Federal Express, was a 1990
Malcolm Baldrige National Quality Award winner.  It released overall
information on customer satisfaction, and we reported in 1992 that 94
percent of Federal Express' customers contacted were completely
satisfied with the overall service. 

Unlike its competitors, however, the 1970 act established the Postal
Service as an executive branch establishment accountable to Congress. 
Since that time, it has increasingly functioned as a businesslike
entity, competing with technology and private companies to deliver
certain services in a competitive marketplace.  However, the duel
objectives of operating as both a public and private entity require
that the Service balance the protection of its competitive interests
with the potential value to the Service and Congress of using the
data, with appropriate safeguards, to help assess and improve
customer service.  We did not review the business customer
satisfaction data compiled by Gallup, but such data would seem useful
to Postal Service management for improving customer service and to
Congress for its oversight activities.  Both the Service and Congress
have found similar data on residential customer satisfaction useful. 

By not receiving any data on business customer satisfaction from the
contractor, the Service and its customers are denied potential
benefits of the Service using the data to improve customer service. 
The data are accumulated by Gallup at considerable cost (projected at
$11.9 million over 4 years).  Meanwhile, as discussed in chapter 3,
the Service is developing plans and has begun numerous national and
local service improvement initiatives.  This is being done without
analyzing and using BCSI results to identify the aspects of service
and the geographic areas indicating the greatest business customer
dissatisfaction.  Disseminating BCSI results to postal management and
providing some of the results, with appropriate safeguards, to
Congress would appear to require little additional cost. 

The limited release of some customer satisfaction data to Congress,
such as was done earlier for residential customers, would not seem to
harm the Service's commercial interests.  Given its experience with
the external distribution of residential data, it appears that the
Service may be able to similarly share some BCSI results with
Congress.  This could perhaps be done by presenting indicators of
business customer satisfaction nationally, for broad customer
groupings, and/or for larger geographic areas.  Where it is
determined that release of the data might hurt the Service, the data
could be made available to appropriate congressional Committees using
appropriate safeguards, such as an agreement with the Committee not
to release the data to the public because it could jeopardize the
Service's commercial interests. 

Congressional oversight Committees for the Postal Service could use
BCSI and other performance data for a variety of purposes, including
ongoing postal oversight activities and consideration of changes to
laws and regulations affecting the Service's performance.  In this
regard, the Postmaster General has said that changes are needed in
aspects of the legislative and regulatory framework that constrain
the Service in pricing its services, introducing new products, and
managing its employees.  Further, legislative proposals are now
pending in Congress to fundamentally change the Service's
governmental status and its responsibilities relating to universal
mail service. 


--------------------
\5 U.S.  Postal Service:  Tracking Customer Satisfaction in a
Competitive Environment (GAO/GGD-93-4, Nov.  12, 1992). 


      BCSI DISTRIBUTION UNDER
      STUDY
-------------------------------------------------------- Chapter 2:4.4

Concerning its plans to distribute BCSI data, Service officials told
us that an officer-level team had been chartered to develop an
overall plan and recommendation for the deployment of both internal
and external measurements used to determine customer satisfaction and
improve customer service.  As part of this effort, the Postal Service
said that it would determine the most effective disposition of the
BCSI.  No date was provided on when the effort would be completed and
whether any BCSI results would be disseminated within the Service or
provided to Congress.  Consequently, the Service did not have a plan
and timetable for using business customer satisfaction data
internally, disseminating the data as appropriate to congressional
oversight Committees, and designing safeguards to protect against the
improper release of sensitive data to competitors. 

In commenting on a draft of this report, the Vice President and
Consumer Advocate said that plans were under way to identify which
managers will receive BCSI results and how frequently they will be
distributed.  She also said that the competitive nature of this
information requires that great care be exercised in making the
information dissemination decisions. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 2:5

The Postal Service's residential customer surveys have provided
valuable data for potential use within the Service and by Congress. 
Postal leadership, particularly the Consumer Advocate, has made
significant progress in disseminating CSI results within the Service
and promoting greater CSI use.  However, opportunities exist to
improve the dissemination and use of CSI results and mail delivery
performance data. 

Perhaps most important is the need for postal management and Congress
to have some indication of how well business customers perceive the
quality of mail service because these customers represent the vast
majority of the Service's business.  Postal leadership is developing
plans, allocating resources, and implementing new service initiatives
without analyzing and using business satisfaction data.  By not using
both business and residential customer satisfaction data, management
attention and resources could be directed disproportionately at
improving those processes that are not of the greatest importance to
overall customer satisfaction and, ultimately, the Service's success. 

It is not reasonable to expect the Service to disclose data on
specific aspects of its services or particular geographic areas that
could jeopardize its competitive interests.  However, the Service's
divergent roles as both a public entity and a business dictate that
it strike a better balance between (1) obtaining and using business
customer satisfaction data to identify and respond to areas of
customer dissatisfaction and providing information to Congress and
(2) protecting business interests by safeguarding against the release
of sensitive, proprietary information.  More general measures of
business and residential customer satisfaction, along with other
performance data such as EXFC ratings, can provide useful yardsticks
for Congress to use in its routine oversight activities and
consideration of legislative proposals that relate to the Postal
Service.  Such data are already compiled and with appropriate
safeguards could be included in the reports that the Service files
annually with oversight and appropriation Committees. 

Because of the Postal Service's investment in national CSI surveys
and the importance of the results to its overall service improvement
efforts, it is important that field offices know both the strengths
and limitations of CSI results and are committed to using the results
as intended by postal headquarters.  CSI reports generated by the
contractor can more fully disclose the level of precision and
usefulness of data.  Users of the reports need to be aware of the
different levels of precision to avoid reaching unwarranted
conclusions, particularly when comparing one organizational component
or geographic area with another or making comparisons over time. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 2:6

To improve the dissemination and potential use of CSI data, we
recommend that the Postmaster General take the following steps: 

  -- As part of the Service's ongoing performance data study,
     establish a plan, safeguards, and timetable for distributing
     business customer satisfaction results to all appropriate
     management levels of the Postal Service for use in improving
     customer service. 

  -- Consult with appropriate congressional committees to determine
     what analyses of business and residential CSI data and other
     available performance data would be useful to them and, using
     appropriate safeguards, provide that data in periodic reports to
     and other communications with Congress for its use. 

  -- Provide more information in the detailed internal CSI reports
     provided by the contractor, including the sampling errors for
     CSI questions and explanations to users on the level of
     precision and usefulness of customer data on certain questions. 


   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 2:7

The Service said that our report presents a generally accurate
picture of what the Service was doing to measure customer
satisfaction and delivery performance at the time of our review and
how the Service could better use the resulting data to improve
service quality. 

The Service did not comment specifically on each of our
recommendations but rather said that it had recently undertaken an
extensive, systematic review of all of its functions and processes. 
The Service said that based on criteria and guidelines of the Malcolm
Baldrige National Quality Award (which we discuss in chapter 3), the
review helped the Service to identify and organize actions necessary
to make its goal of a customer-driven, customer-oriented, and
customer-responsive organization a reality.  The assessment led to a
program the Service calls CustomerPerfect!sm

The Service said that our recommendations and concerns regarding
information sharing will be addressed in that program.  For example,
the Service said that a team headed by the Consumer Advocate was
studying the dissemination of customer satisfaction results, both for
business and residential customers.  The team was to develop a
strategy for making survey results available to the public and
Congress. 

The Service expressed some minor disagreement regarding our
comparison of CSI and EXFC ratings.  The Service inferred that we
anticipated more of a connection between the CSI and EXFC ratings
than the data actually show and explained that the two ratings are
different.  We agree that CSI and EXFC are very different measures,
and we had no preconceived notion that the results of the CSI surveys
would "match" or closely relate to measures of on-time delivery
performance under EXFC.  Rather, our purpose was to show how the
ratings differ and emphasize how the Service could explain the extent
and reasons for differences in customer perception of the Service's
performance, as measured in CSI, and its delivery performance, as
measured in EXFC.  To help clarify this point, we made some changes
to the section of the report comparing CSI and EXFC ratings. 

The Service also said that its customer satisfaction and delivery
systems are useful as measurement tools but less useful for
diagnostic purposes.  The Service wants to improve the systems to
provide more precise and immediate feedback for making real-time
improvements in service quality. 


POSTAL HEADQUARTERS CAN STRENGTHEN
ITS OVERALL PLANNING AND
MONITORING OF SERVICE IMPROVEMENT
INITIATIVES
============================================================ Chapter 3

The Service has had many innovative and promising service improvement
efforts under way but, as discussed in chapter 2, it has not used
business customer satisfaction data as part of these efforts. 
Available residential CSI data show that the level of customer
satisfaction remained about the same in 1994 as in 1991.  Despite its
many initiatives, most of which began in 1990 and 1991, the Service
has not implemented at the performance cluster level a corporatewide
strategy for improving customer satisfaction and focusing all field
offices on the most significant underlying cause of customer
dissatisfaction; namely, unreliable mail delivery.  The Service's
performance incentive plans for managers and employees did not
include available measures of delivery service reliability, such as
EXFC data.  Further, postal headquarters did not follow a systematic
approach for (1) monitoring field offices' progress in improvement
initiatives and (2) sharing information among all field offices on
the best customer service practices. 


   NUMEROUS NATIONAL AND LOCAL
   INITIATIVES UNDERTAKEN TO
   IMPROVE CUSTOMER SERVICE
---------------------------------------------------------- Chapter 3:1

Using residential CSI data and other performance indicators, the
Postal Service has begun numerous efforts to improve customer service
and reduce significant levels of customer dissatisfaction.  These
efforts have included (1) encouraging, training, and rewarding
employees to better serve customers; and (2) setting new policies and
standards to focus greater corporatewide attention on customer
service.  In line with national customer service goals, field offices
have pursued a broad array of efforts to improve service. 


      NATIONAL INITIATIVES ARE
      EMPLOYEE AND CUSTOMER
      FOCUSED
-------------------------------------------------------- Chapter 3:1.1

The Service's employee-related efforts are designed to better
recognize the importance of postal employees to substantial and
sustained improvements to customer satisfaction.  The influence of
postal employees on customer satisfaction can be seen in CSI results. 
Customers indicate in the residential CSI surveys whether they have
visited, phoned, or complained to their local post offices during the
quarter covered by each survey.  Analyses of CSI data done by the
Office of Consumer Advocate and the Postal Inspection Service show
that the more contact a customer had with the Postal Service, the
lower the customer rated its overall performance.  For example, the
Inspection Service reported in December 1994 that customers who had
not gone into a post office in the 3 months preceding the CSI survey
gave the Service higher marks than those who had visited a post
office during the same period. 

The Postal Service, acting unilaterally in some cases and in
cooperation with the unions in other cases, has taken numerous steps
to stimulate greater employee commitment to serving customers.  Its
initiatives since 1990 include the following. 

  -- Employee opinion surveys (EOS) are done annually to obtain and
     track over time the views of employees at all organizational
     levels regarding their jobs, the organization, customers, and
     other topics. 

  -- New incentive payment plans were implemented to base employee
     rewards, in part, on the Postal Service's performance in
     improving customer satisfaction and meeting financial goals. 

  -- As part of a corporatewide "Quality First!" initiative, training
     was provided to thousands of headquarters and field office
     employees to promote a total quality approach uniformly
     throughout the Postal Service.  Subsequently, in lieu of the
     Quality First!  initiative, the Service adopted the Malcolm
     Baldrige National Quality Award criteria\1 for encouraging,
     facilitating, and measuring the Postal Service's commitment to
     improving customer satisfaction. 

  -- Courtesy and sales training was provided for both craft and
     management employees involved in retail operations to improve
     skills and the motivation that leads to greater customer
     satisfaction and revenue generation. 


--------------------
\1 Named for former Secretary of Commerce Malcolm Baldrige, this
award is presented annually to organizations that have demonstrated
success in adopting quality principles, as measured against various
performance criteria generally available to organizations. 


      NEW POLICIES AND STANDARDS
      ARE MORE CUSTOMER-ORIENTED
-------------------------------------------------------- Chapter 3:1.2

The Service has also adopted new corporatewide retail policies and
standards and, at the time of our review, it was acquiring new retail
equipment and facilities to improve responsiveness to customer needs
and expectations.  These efforts, begun or expanded since 1990,
include the following. 

  -- The Service increased customer convenience by expanding an "Easy
     Stamp" program to allow customers to buy stamps by phone, mail,
     a computer network, and automatic teller machines. 

  -- Debit and credit cards are accepted for the purchase of stamps
     and certain other transactions. 

  -- A national standard of "Service In Five Minutes or Less" was
     adopted to reduce customers' waiting time in line at some post
     offices, and post office hours were adjusted to better meet
     customer needs. 

  -- New retail service equipment was acquired, such as stamp vending
     machines, terminals for use by window clerks, and postage
     validation machines. 

  -- A new postal retail store design was approved for post offices
     to be constructed or renovated, and a new design for lobbies in
     some existing post offices was also approved.  Both efforts were
     intended to provide interior appearances more appealing to
     customers than traditional post offices and make services more
     readily accessible. 

  -- Customer advisory councils were formed to solicit customer
     feedback from local community residents. 

  -- "Customer care centers" were established to help improve receipt
     and handling of customer calls, and a 1-800 toll-free service
     was set up for resolving the complaints of customers who
     continued to have problems after contacting local post offices. 


      SOME FIELD OFFICES WERE
      PURSUING PROMISING SERVICE
      IMPROVEMENT EFFORTS
-------------------------------------------------------- Chapter 3:1.3

Our review, and two related reviews done by the Postal Inspection
Service, revealed a wide array of imaginative and potentially
successful efforts under way at some field offices.  Following are
brief summaries of some efforts that were under way in one or more of
the six districts that we visited. 


      BOX CALL PROJECT
-------------------------------------------------------- Chapter 3:1.4

The Postmaster of Springfield, MA, undertook a box call project at
his main post office to enable post office box customers to call a
central phone number to determine whether they had mail in their
boxes.  The basic premise for the project was that customers would
appreciate saving a trip to the post office if they did not have
mail.  Post office employees input into a hand-held device the box
numbers that have no mail.  These numbers are then downloaded into a
personal computer.  Customers access the computer by telephone, key
in their box numbers, and are told whether they have mail. 


      PARCEL LOCKERS IN APARTMENT
      BUILDINGS
-------------------------------------------------------- Chapter 3:1.5

The New York, NY, District initiated a program to place parcel
lockers in apartment buildings.  When a tenant who is not at home
receives a parcel, the parcel is put into one of the lockers and the
key to the locker is put into the tenant's mail box.  After the
tenant inserts the key into the parcel box to retrieve the package,
the key has to be removed with the carrier's master key. 

One program objective was to serve senior citizens and disabled
persons who may have difficulty getting to the post office or
carrying heavy packages back to their homes.  It also allows other
customers who are not home during the day to obtain parcels
conveniently without delay, and without having to wait in post office
window service lines. 

Criteria the Service used in deciding whether to install a parcel
locker inside an apartment house included the number of undelivered
packages on a carrier route and how far the post office was from the
apartment building.  As of August 1993, the District had more than
2,000 parcel boxes in 172 high-rise buildings in Manhattan. 


      MYSTERY CALLER PROGRAM
-------------------------------------------------------- Chapter 3:1.6

Recognizing the need to improve CSI scores for telephone assistance,
the New York, NY, District established a Mystery Caller Program in
1993 under its Customer Services Support group.  The aim was to
ensure quicker response time, improve the accuracy of answers to
customers, and improve clerk courtesy.  During each 2-week period, 4
calls were placed to each of the district's 117 stations.  To achieve
satisfactory performance, a station must score 26 out of 32 possible
points, or a score of 80 percent.  The program was nicknamed the "100
Club" to encourage stations to respond enthusiastically and to seek a
perfect score.  Any station receiving a perfect score of 100 for six
consecutive rating periods receives a bronze plaque recognizing that
accomplishment.  Silver, gold, and platinum plaques are presented
when a station receives a perfect score for two, three, and four
consecutive rating periods, respectively.  Plaques are displayed in
the station lobby.  On the other hand, a station receiving a score of
80 or less must submit a plan to improve its rating. 


      REDUCING NO-RECORD MAIL
-------------------------------------------------------- Chapter 3:1.7

Retail units can accept change of address notices from customers who
move so that First-Class mail arriving at their former addresses is
forwarded to them for a period of 12 months.  The Postal Service
employs computerized forwarding sites (CFS) for keeping track of
forwarding addresses and applying new address labels to mail to be
forwarded.  However, if a post office sends mail to a CFS for
forwarding but the CFS finds no forwarding data in the computer, the
CFS returns the mail to the post office as "no-record" mail.  Mail is
frequently returned when the customer's mail forwarding date has
expired. 

In February 1993, the Bellevue, NE, (Central Plains District) post
office had a no-record rate of 17 percent (5 percent or below is
considered good).  The Bellevue Postmaster agreed with the post
office operations manager to whom he reports to reduce the post
office's rate.  A task force composed of management and craft
employees was set up to work toward reducing the rate of no-record
mail.  The task force met to establish project objectives, develop an
action plan, and set time frames.  Subsequently, the task force
visited the CFS to observe the processing of Bellevue's CFS mail. 
The CFS processed Bellevue's no-record mail during the visit and
sorted it by carrier route so the task force could speak with the
individual carrier about his or her mail.  Later, the postmaster
developed a procedure to double-check mail before it is sent to the
CFS. 

The task force evaluated the project through weekly reports from the
CFS.  It also planned an ongoing dialog with the CFS supervisor to
correct future problems.  As of June 1993, the no-record mail
percentage at Bellevue had been reduced to the Postmaster's goal of
about 10 percent.  Appendix V provides information on other
improvement initiatives that we identified in the six districts that
we visited. 


   SIGNIFICANT LEVELS OF
   RESIDENTIAL CUSTOMER
   DISSATISFACTION HAVE CONTINUED
   SINCE 1991
---------------------------------------------------------- Chapter 3:2

Many of the Service's improvement initiatives were still being
implemented at the time of our review.  Further, the Service believes
that many more years of concentrated effort at all levels of the
organization will be required before breakthrough improvements in
customer satisfaction can be expected.  The Service's measures of
residential customer satisfaction and its delivery performance
support this notion. 

Through 1994, residential CSI data and other performance data show
that the Postal Service is having little sustained success in its
efforts to reduce customer dissatisfaction by improving customer
service.  In November 1993, the Postmaster General announced a
favorable CSI rating (excellent, very good, and good responses) of 89
percent--the highest ever achieved.  He said that actions were under
way to improve that rating by 2 percentage points.  Since that time,
however, the favorable rating dropped to 85 percent in postal quarter
4, 1994 (May 28, 1994, to September 16, 1994).  This was the same
rating reported for the first quarter in which all 170 metropolitan
areas were measured by CSI in 1991.  During the 14 postal quarters
through September 1994, the favorable ratings ranged from 85 to 89
percent nationally, and the excellent and very good ratings ranged
from 51 to 60 percent, with a rating of 52 percent reported for
postal quarter 4, 1994.  (See fig.  3.1.)

   Figure 3.1:  Customer
   Satisfaction Ratings Were About
   the Same at the End of 1994 and
   1991

   (See figure in printed
   edition.)

Source:  CSI Surveys, U.S.  Postal Service. 

The 85 percent CSI rating meant that on the basis of the Service's
survey of a representative sample of about 90 million households,
13.5 million households rated the Postal Service's performance as
fair or poor--an increase of about 3.6 million from a year earlier. 
The national CSI ratings for postal quarters 1, 2, and 3, 1995, were
85 percent, 85 percent, and 86 percent, respectively.  National CSI
ratings differ from those for many metropolitan areas.  The ratings
for quarter 4, 1994, for some metropolitan areas were up to 6
percentage points above the national average.  For some other areas,
the ratings were as much as 34 percentage points below the national
average. 

As discussed in chapter 2, neither we nor postal management have
access to similar data on levels and trends of business customer
satisfaction gathered by an independent contractor.  As a result, we
could not determine whether the satisfaction of these customers is
better or worse than residential customers and if business customer
satisfaction has improved since it was first measured in early 1994. 

EXFC data show that the national rating for on-time delivery has yet
to exceed 90 percent, even though the Service's goal is to deliver
all First-Class mail on time 95 percent of the time.  As indicated in
figure 3.2, EXFC ratings have ranged from 79 to 84 percent nationally
for the 14 quarters ended September 1994, with a rating of 83 percent
reported for postal quarter 4, 1994. 

   Figure 3.2:  External
   First-Class On-Time Delivery
   Ratings Have Ranged From 79 to
   84 Percent During 1991 Through
   1994

   (See figure in printed
   edition.)

Source:  EXFC System, U.S.  Postal Service. 

The national EXFC rating was 87 percent for postal quarter 4, 1995,
ending in May 1995.  This was the highest national EXFC rating ever
reported by the Postal Service. 

We also reviewed CSI and EXFC data to determine the number of
metropolitan areas that had higher and lower ratings during postal
quarter 4, 1994, compared with the same period 3 years earlier in
fiscal year 1991.  We identified those metropolitan areas with CSI
and EXFC changes of more than 3 percentage points during this period
because percentage changes of less than this could be due to chance. 
For most metropolitan areas, customer satisfaction and on-time
delivery performance dropped.  Specifically, CSI ratings dropped for
20 of 31 metropolitan areas and increased by more than 3 percentage
points for the remaining 11 during the 3-year period.  EXFC ratings
dropped for 27 of 43 areas and increased for 16. 


      IMPROVED CUSTOMER
      SATISFACTION IS IMPORTANT TO
      CUSTOMER RETENTION
-------------------------------------------------------- Chapter 3:2.1

The Postal Service wants to reduce customer dissatisfaction because
it is concerned that customers will shift to other suppliers of mail
services, where they are available or should they become available. 
Postal Service data indicate that many customers would use other
services if given the opportunity.  According to the Postal
Inspection Service's analysis of CSI data, 42 percent of all
residential customers would be at risk of shifting to another service
if another service were available.\2

The Inspection Service's analysis was based on responses to the
following CSI question: 

     "Right now the only way to mail a First-Class letter is through
     the U.S.  Postal Service, but if there were another mail service
     which you could use to mail a letter at the same price, would
     you switch to another service?"

According to the Inspection Service report, many customers who rated
the Service's overall performance as excellent, very good, and good
are at risk of shifting to another service.  Postal Service
management officials also said that many residential customers might
switch to another service.  They said that over 40 percent of the
residential customer market is vulnerable to competition from another
service, assuming that the postage charged is the same as that of the
Postal Service. 


--------------------
\2 As previously stated in chapter 2, the Private Express Statutes
place restrictions on customers' use of services other than the U.S. 
Postal Service for delivery of letters. 


   NO OVERALL STRATEGY YET
   IMPLEMENTED TO FOCUS ALL FIELD
   OFFICES ON KEY CAUSES OF
   CUSTOMER DISSATISFACTION
---------------------------------------------------------- Chapter 3:3

Our review, and related reviews done by the Postal Inspection
Service, show that the use of CSI data and the development of related
improvement initiatives have not followed an overall national
strategy for focusing field offices' attention on the principal
causes of customer dissatisfaction.  To a large extent, the
improvement efforts initiated on the basis of CSI results have
focused on post office operations, such as window and lobby services. 
The efforts did not always encompass employees and operations in mail
processing plants or focus on specific aspects of service, such as
the consistency and reliability of mail delivery, that CSI results
indicate offer the greatest opportunity to improve customer
satisfaction. 


      IMPROVING RELIABILITY OF
      DELIVERY SERVICE OFFERS
      GREATEST OPPORTUNITY TO
      IMPROVE CUSTOMER
      SATISFACTION
-------------------------------------------------------- Chapter 3:3.1

Residential CSI data can be analyzed to identify aspects of service
causing the greatest customer dissatisfaction.  Such analyses show
that improving the reliability (i.e., on-time delivery rates) of mail
service offers the greatest potential for the Postal Service to
improve customer satisfaction.  Each quarter, detailed CSI reports
prepared by the contractors rank responses to 37 questions in terms
of their relative importance as "drivers" of customer satisfaction. 
The rankings represent the level of improvement potential calculated
on the basis of the number of customers responding to each question
and the number of good, fair, and poor responses to each. 

Of the 37 questions on specific aspects of service, those on the
reliability of delivery time for local and nonlocal mail represented
the greatest opportunity for the Postal Service to improve customer
satisfaction.  Those aspects of service that offer the least
potential for improvement are under the control of postmasters and
include window and lobby services offered at post offices, mail
forwarding, and telephone service. 

While the Service has not obtained and analyzed BCSI data, as
discussed in chapter 2, other data show that reliable delivery
service is of greatest importance to all of the Service's customers,
both business and residential.  According to Consumer Advocate data,
customers complained more about late and missent mail than any other
aspect of the Postal Service's performance in fiscal year 1994 and
other recent years.  Moreover, we previously reported\3 that the
Postal Service has lost overnight and parcel delivery service,
involving primarily business customers, to competitors, in part
because those competitors offered faster and more reliable delivery. 


--------------------
\3 U.S.  Postal Service:  Pricing Postal Services in a Competitive
Environment (GAO/GGD-92-49, March 25, 1992). 


      PROCESSING PLANTS MINIMALLY
      INVOLVED IN USING CSI
      RESULTS TO DEVELOP
      IMPROVEMENT PROJECTS
-------------------------------------------------------- Chapter 3:3.2

How mail processing plants operate can significantly affect customer
satisfaction, but the plants have been less involved than customer
service districts and post offices in using CSI data to improve
customer satisfaction.  The work done at processing plants can have a
major influence on the reliability of mail delivery, the most
important aspect of service to customer satisfaction. 

Postal Inspection Service reports on CSI issued in September 1992 and
December 1994 showed that postal management had tended to focus
improvement initiatives on processes and employees in customer
service districts.  In its December 1994 report, the Inspection
Service reported that while all aspects of customer service require
continuing attention, processing plants continued to be minimally
involved in analyzing CSI data and planning and implementing
activities to increase customer satisfaction.  Customer service
districts had taken the lead in using CSI data, and their actions
generally included only post office and carrier services and not the
operations at processing plants.  The Inspection Service also
reported that the districts tended to direct efforts at "quick fix"
categories of CSI questions, such as complaint handling and telephone
service, that have relatively low potential for improving customer
service. 

Managers in mail processing did, however, use EXFC data to emphasize
timely processing of mail.  They also used other performance
indicators, such as volumes of mail left at plants at the end of
processing cycles. 

Some of our earlier reviews showed that the Postal Service's
principal improvement initiative in processing plants has been the
automation of mail sorting, which began in 1982.  In 1993, the
Service began to automate the sorting of letter mail to each home and
business address to relieve carriers of this workload.  The Service's
automation goal has primarily been to reduce work hours and
employees, not to improve delivery service by reducing mail cycle
times.  However, in December 1994, Postal Service officials did
report to the Board of Governors for the first time that certain
barcoded mail, which can be sorted automatically, was delivered
faster to customers than nonbarcoded letters.\4

Postal Service employee opinion surveys show that clerks, mail
handlers, and other mail processing employees are less aware than
employees in customer service districts of how their work affects
customer satisfaction.  However, the work that mail processing
employees do affects the reliability of mail delivery and, hence,
customer satisfaction.  These employees must receive, sort, and
dispatch mail according to preestablished schedules for the Postal
Service to meet delivery dates promised to customers.  To emphasize
the importance of the role played by mail processing employees in
customer satisfaction, the inspectors' report included the following
example: 

     "A mailhandler pulling a container of trayed mail to the dock
     for dispatch was asked how he affected customer satisfaction. 
     He replied he doesn't see or deal with customers.  It was
     pointed out to him if a carrier makes a misdelivery, that
     carrier has affected one, maybe two customers, but if a mail
     handler places a container of mail on the wrong truck, he may
     affect 50,000 customers in a detrimental way."

Several of the Service's initiatives, such as the EOS and Quality
First!  initiatives, did encompass managers and employees in mail
processing plants.  In providing Quality First!  training, the
Service instructed field employees on use of CSI and EXFC data in
improving the reliability of mail service. 


--------------------
\4 For more information on the Service's automation program, see
Postal Service:  Automation Is Taking Longer and Producing Less Than
Expected (GAO/GGD-95-89BR, Feb.  22, 1995). 


      UNIONS AND MANAGEMENT HAVE
      GENERALLY BEEN UNABLE TO
      AGREE ON EMPLOYEE
      INVOLVEMENT
-------------------------------------------------------- Chapter 3:3.3

Most of the employees at mail processing plants are members of the
American Postal Workers Union (APWU), which in the past has not
participated in the Service's initiatives to involve employees in
service improvement efforts.  In this regard, we recently reported\5
that breakthrough improvements in customer service cannot be achieved
unless the Postal Service and labor unions representing postal
employees resolve long-standing workfloor problems.  Postal
management has had difficulty getting labor unions to agree on the
involvement of employees with each other and with management in
solving customer service and other problems.  For example, APWU is
the largest postal union and did not participate in initiatives, such
as Striving for Excellence Together (SET), Employee Involvement, and
Quality of Work Life, which are described in our earlier reports. 
Neither APWU nor the National Association of Letter Carriers, which
together represent about 85 percent of the total number of craft
employees, participate in the SET program. 

As we reported earlier, a lack of labor-management cooperation has
been a serious limitation on the Service's ability to make
significant, sustained improvements in customer satisfaction.  As of
July 1995, the Service and three of its four major unions (the rural
letter carrier union being the exception) had not agreed to meet and
begin developing new approaches to involve employees with union and
management leaders in improving the processing and delivery functions
of the Postal Service. 


--------------------
\5 D.C.  Area Mail Delivery Service :  Resolving Labor Relations and
Operational Problems Key to Service Improvement (GAO/GGD-95-77, Feb. 
23, 1995); U.S.  Postal Service:  Labor-Management Problems Persist
on the Workroom Floor (GAO/GGD-94-201 A and B, Sept.  29, 1994). 


      OVERALL PLAN WAS TO BE
      DEVELOPED TO GUIDE CSI
      IMPROVEMENT EFFORTS OF ALL
      FIELD OFFICES
-------------------------------------------------------- Chapter 3:3.4

Although it had numerous improvement efforts under way, the Postal
Service did not have at the time of our review an overall plan to
guide and integrate all of its CSI-related improvement efforts at
post offices and processing plants.  During our review, the Postal
Inspection Service issued its December 1994 report and recommended
that the Service develop a plan involving all field offices in the
use of residential CSI data to improve customer satisfaction. 

In response to that recommendation, the Vice President for Work Force
Planning and Service Management said that a corporate service plan
would be developed, with emphasis on the role of processing and
distribution as well as customer service in jointly improving service
levels, as measured by CSI and other systems.  The development of the
plan was to begin in January 1995, and implementation was to begin
within 120 days after the plan was finalized. 

We did some follow-up after completing our field work to determine
the status of the plan and were told that some effort had been made
to develop a plan.  This included identifying 108 separate
headquarters service improvement efforts, relating to business and/or
residential customers, under way in early 1995. 

However, this effort to integrate all of the Service's CSI-related
initiatives was discontinued as such.  In June 1995, the Vice
President responsible for developing the plan advised the Postal
Inspection Service that major changes had occurred in the corporate
approach to improving customer satisfaction.  One such change was a
decision to apply the Malcolm Baldrige National Quality Award
Criteria to the Postal Service, mentioned earlier. 

According to Service officials, the Baldrige initiative was started
in 1994, with the guidance of a new Vice President for Quality and
outside consulting services.  In this initiative, the Service had set
up 10 teams, including a team of senior leaders headed by the
Postmaster General and an information and analysis team headed by the
Vice President for Work Force Planning and Service Management.  The
functions of these teams included those described in seven categories
of Baldrige criteria.\6 As a first step, an outside consulting firm
worked with the 10 area offices and 10 of the 85 performance clusters
to assess current conditions against the criteria and provide
baseline data for future assessments.  In March 1995, the 10 teams
were created and began developing actions plans for applying the
criteria to the particular deficiencies identified by the Baldrige
assessment. 

The Service's plan to apply the Baldrige criteria, in what it refers
to as CustomerPerfect!, appears to be another innovative and
promising initiative that could make a difference in future levels of
customer satisfaction.  However, as with some of the Service's past
initiatives, labor unions representing postal employees are not a
part of this new initiative.  According to the Vice President for
Work Force Planning and Service Management, leaders of major unions
discontinued their participation in meetings with postal leadership
when contract negotiations began in August 1994.  He said that the
unions are not represented on any of the 10 teams set up to implement
the Baldrige criteria. 


--------------------
\6 Malcolm Baldrige National Quality Award - 1995 Award Criteria
published by the National Institute of Standards and Technology,
Department of Commerce, list the seven categories of criteria as
follows:  leadership, information and analysis, strategic planning,
human resource development and management, process management,
business results, and customer focus and satisfaction. 


   EMPLOYEE PAY INCENTIVE PLANS
   INCLUDE CSI RESULTS BUT NOT
   SOME KEY MEASURES OF SERVICE
   RELIABILITY
---------------------------------------------------------- Chapter 3:4

Along with developing an overall plan and pursuing other service
improvement efforts, the Postal Service has continued to reward
certain employees for their performance partly on the basis of CSI
results.  These performance incentives, which we consider an
innovative approach to linking employee pay more closely to
organizational performance, are used to focus greater management and
employee attention on customer service.\7 The incentive payments are
based on residential customers' perceptions of the Service's overall
performance and other measures relating to financial performance and
employee relations.  However, it seems to us that the plans may be
more effective if they also incorporate some of the key measures of
service reliability, such as EXFC delivery performance data. 
Moreover, because the Service has not yet obtained and used BCSI
results, the incentive plans do not incorporate levels and changes of
satisfaction among the business customers, representing about 90
percent of the Service's business. 

Some craft employees, all supervisors and managers, and most
executives are all rewarded, in part, on the basis of CSI results. 
As part of union contract negotiations in 1990, the Service and 2
unions agreed to use 2 factors, CSI results by performance cluster
and Service-wide financial (budget) performance, to make annual
performance incentive payments to certain craft employees (92,852
employees, or about 15 percent of the craft work force in 1994) under
the SET program. 

Subsequently, in consultation with the management associations,\8

the Service extended the performance incentive plan to all
supervisors and managers, i.e., those covered by the Service's
Executive and Administrative Schedule (EAS).  In addition, the
Service later began to base incentive payments to most executives
(about 950) in the Postal Career Executive Service (PCES)\9 on local
and national CSI results, financial performance, and EOS results. 

Customers' perception of the Service's overall performance, as
indicated by responses to question 1a in the residential CSI surveys
and reported by performance cluster, dictate the difference in
incentive payments to those craft employees, supervisors, and
managers covered by the incentive plans.  Payments based on financial
performance are the same for all of these employees.  Incentive
payments to executives vary depending on question 1a results for
groups of performance clusters or nationally and national financial
performance.  In fiscal year 1994, the Service incorporated EOS
survey results into the incentive program for PCES-I employees. 
(App.  IV provides additional details on the incentive pay plans.)

None of the incentive plans include available EXFC data and other
available delivery measures (e.g., measures of second- and
third-class on-time delivery).  We believe that the recognition of
such delivery measures in the incentive plans is important because,
as discussed previously, CSI data analysis shows that improving
service reliability offers the greatest opportunity for improving
customer satisfaction.  Moreover, CSI and EXFC data show that a wide
gap often exists between customers' perceptions of the Service's
performance and its actual delivery performance.  Consequently, in
using only the overall CSI rating as one component in determining
performance payments, the Service rewards employees on the basis of
factors that are less under their control, i.e., perceptions of the
Postal Service, than some other factors that are more under their
control, i.e., mail collection, transportation, sorting, and
delivery. 

If CSI and EXFC ratings generally were consistent with each other,
the use of CSI alone would not be nearly as consequential.  However,
the Service's delivery performance often differs significantly from
customers' perceptions of its overall performance for many
metropolitan areas. 

We visited the Springfield, MA, and Chicago, IL, metropolitan areas
because the former was among the highest-ranking of all clusters in
CSI ratings and the latter was among the lowest.  We found that
management in both areas were using CSI results to emphasize the need
to improve customer service and had a number of initiatives under way
to improve service.  Moreover, the differences in EXFC ratings for
the two areas (shown in figure 3.3) were much smaller than
differences in their CSI ratings (shown in figure 3.4). 

   Figure 3.3:  EXFC Ratings For
   Springfield, MA, and Chicago,
   IL, for 13 Postal Quarters
   Ended September 1994

   (See figure in printed
   edition.)

Source:  EXFC System, U.S.  Postal Service. 

   Figure 3.4:  CSI Ratings for
   Springfield, MA, and Chicago,
   IL, for 13 Postal Quarters,
   Ending September 1994

   (See figure in printed
   edition.)

Source:  CSI Surveys, U.S.  Postal Service. 

Our analysis showed that the relationship of EXFC and CSI ratings for
some other metropolitan areas was similar to the above two areas. 
Further, many metropolitan areas having the highest CSI ratings in
1994 also had similarly high ratings at the time of the first CSI
survey in 1991, before the Service began many of its current
improvement initiatives.  In contrast, some areas having the lowest
ratings experienced significant change in CSI scores over the same
period, as figures 3.5 and 3.6 show for selected high-ranking and
low-ranking areas. 

   Figure 3.5:  Top Four
   Metropolitan Areas Had
   Relatively Little Change in
   Overall Satisfaction From
   Fiscal Year 1992 Through 1994

   (See figure in printed
   edition.)

Note:  CSI percentages include excellent, very good, and good
ratings. 

Source:  CSI Surveys, U.S.  Postal Service. 

   Figure 3.6:  Change in Bottom
   Four Metropolitan Areas in
   Overall Satisfaction From
   Fiscal Year 1992 Through 1994

   (See figure in printed
   edition.)

Source:  CSI Surveys, U.S.  Postal Service. 

Customer perception of the Service's overall performance is only one
indicator of its performance and service quality.  Other indicators
that are available to the Postal Service include not only the results
of the Service's independent measurements of First-Class,
second-class, and third-class mail delivery service, but also
customers' responses to some of the specific CSI questions.  CSI
surveys include questions that relate directly to delivery
performance; customers are asked about their satisfaction with both
local and nonlocal mail delivery. 

Including the Service's available delivery performance measures,
broken out by performance cluster, in the calculation of incentive
payments would appear to provide a more direct link between the
incentive payments and both (1) mail delivery processes that are most
under performance cluster employees' control and (2) the factors that
are most important to customer satisfaction. 

Postal officials administering the incentive plan said that the
incentive payments for some employees were developed as a result of
1990 contract negotiations with unions, before the first CSI and EXFC
results were available.  These officials said that the objectives at
that time were to get the plan adopted and to keep it simple.  The
Service's Vice President for Quality, who came to the Postal Service
in July 1994, said that he was concerned about the Service's heavy
reliance on customer perception as a single or principal performance
measure.  The Vice President said the Service is reexamining its
collection and use of all externally generated data.  He also said
that several new efforts, including the application of Baldrige
criteria, are under way to focus greater attention on those processes
that employees can improve. 

In September 1995, after reviewing a draft of this report, the Vice
Presidents for Quality and Human Resources said that as part of
CustomerPerfect!, the Service will be aligning the PCES compensation
system for PCES employees.  The Vice President for Quality said that
this new compensation alignment will consider EXFC, CSI, and BCSI
measurements.  He also said that a similar alignment will be proposed
in the next round of consultations with management associations and
negotiations with the unions. 


--------------------
\7 The incentive plans and payments discussed above are team-based
and are in addition to the recognition of executives, managers, and
supervisors under the Service's Exceptional Individual Performance
Program.  This program is designed to recognize a small number of
employees whose performance is extraordinary. 

\8 The Postal Service is required to consult with management
associations on changes affecting postmasters and supervisors.  The
three associations are the National Association of Postal
Supervisors, the National League of Postmasters, and the National
Association of Postmasters of the United States. 

\9 In 1994, the Service had 956 PCES-I executives covered by a
performance incentive plan under which CSI results are one of the
factors used for calculating incentive payments.  These executives
are responsible for national policies or major organizational units. 
There were 34 PCES-II executives not covered by the plan.  These
executives are officers (other than the Postmaster General and the
Deputy Postmaster General) responsible for broad functional
departments and the heads of the 10 postal area offices. 


   SOME NATIONAL INITIATIVES NOT
   SYSTEMATICALLY TRACKED
---------------------------------------------------------- Chapter 3:5

After starting some national improvement initiatives in 1992, postal
headquarters did not regularly follow up to determine the extent to
which the initiatives were implemented and if they improved customer
satisfaction.  Such follow-up would allow headquarters to assess
field offices' progress in implementing national initiatives in a
timely manner and share with other field offices the best practices
of post offices and processing plants in serving customers. 


      CURRENT APPROACH ENCOURAGES
      INNOVATION AND RECOGNIZES
      DIFFERENCES AMONG FIELD
      OFFICES
-------------------------------------------------------- Chapter 3:5.1

The Postal Service followed a decentralized approach to implementing
new initiatives.  Its approach encouraged employees in post offices
and processing plants to be innovative in working together and with
customers to solve service problems.  This approach recognizes that
the field structure is large and complex--hundreds of mail processing
facilities and more than 40,000 post offices, branches, and stations. 
For example, the number of postal employees assigned to the 6
customer service districts that we visited ranged from 2,300 in
Billings, MT, to 10,800 in New York, NY.  (See app.  II for
additional information on the relative size of the six districts.)

Post offices also operate in a variety of environments to meet a
broad array of customer needs.  For example, postal officials in
Billings, MT, had relatively little concern about the security of
postal customers, employees, and equipment, allowing them to provide
convenient access to window and lobby services.  In contrast,
physical security was of great concern to some post offices in the
New York City area.  There, bullet-proof glass protected clerks from
the public, and lobbies were locked after certain hours. 

Without changing its decentralized approach to implementing
improvement initiatives, Postal Service headquarters could use a more
systematic and uniform approach for tracking field offices'
implementation of national initiatives and reporting the impact of
the initiatives on CSI ratings and revenue.  As indicated above,
field offices were pursuing numerous retail initiatives.  The time
projected for completing the initiatives spans many years, and a
number of postal headquarters' offices were overseeing the
initiatives. 


      TRACKING OF FIELD OFFICE
      PROGRESS VARIED AMONG
      HEADQUARTERS OFFICES
-------------------------------------------------------- Chapter 3:5.2

The tracking of national initiatives that we reviewed varied among
headquarters offices, with procedures and data on some initiatives
being more extensive than for others.  The Office of Consumer Affairs
had gathered fairly extensive data for monitoring the status and
results of efforts to improve telephone service.  For example, for
the centralized call centers, the Office had set time standards for
resolving customer complaints and keeping customers informed.  Each
customer complaint was to be logged, and a case history and caller
profile were to be developed so that the complaint could be tracked
until final resolution.  The office's data showed that about 60
percent of complaints received at the centralized call centers
through December 1994 were resolved by employees at the center.  The
remaining calls required assistance from district or post office
employees. 

The Office of Consumer Affairs was also monitoring the use of
Consumer Advisory Councils.  Postmasters were to decide when they
wanted to set up a council, and through December 1994, relatively few
post offices had formed councils.  The first council was established
by the Honolulu, HI, district in 1988, and 16 additional councils had
been formed by the end of that year.  By December 1994, 1,572
councils were operating nationwide. 

Some other headquarters units were still developing procedures to
track the implementation and results of national initiatives under
their responsibility.  For example, a retail support group under the
Vice President for Marketing was responsible for overseeing several
initiatives to be implemented by post offices.  In June 1993, the
group requested area and district offices to provide data on post
offices that had announced the service in 5 minutes or less standard
and that had adjusted window hours.  However, the data provided were
incomplete.  Of the 85 customer service districts, only 56 districts
responded to the request.  The 56 districts reported that of the
approximately 40,000 post offices, branches, and stations nationwide,
about 5,000 post offices had posted the 5 minutes or less service
standard.  The retail support group did not have data to show if
those post offices serving large numbers of customers each day, and
thus possibly having the greatest difficulty providing service, had
announced the standard and adjusted hours.  Postal retail officials
said that no further effort had been made to obtain data on the two
initiatives. 

The group was planning to track changes in CSI ratings as field
offices implemented the 5 minute or less standard and expanded window
hours.  In addition, the group was considering different methods for
determining whether post offices were meeting the service within the
5 minutes or less standard, and the group was considering methods for
measuring the impact on postal revenue of adjusted window hours. 

Headquarters staff were also developing a plan to track the
implementation and results of the new retail store initiative. 
Evaluations were to include customer and employee responses to the
new design as well as revenue analysis.  The evaluations were
expected to include a breakdown of revenue sources (e.g., packaging
products and vending machines) and cost studies of various
implementation approaches, contractor performance, and ease of
implementation. 

Although we obtained information mainly on initiatives of customer
service districts, the recent planning efforts by the Vice President
for Work Force Planning and Service Management discussed earlier
showed that postal headquarters was not systematically tracking other
national customer satisfaction initiatives.  That office did a
one-time survey of ongoing initiatives and identified 108 projects
under way in early 1995.  However, no further steps were taken at
that time to integrate and assess the projects because, as mentioned
previously, that effort was superseded by other broader headquarters
initiatives. 


      PROCEDURES FOR SHARING LOCAL
      INNOVATIONS DISCONTINUED
-------------------------------------------------------- Chapter 3:5.3

Our review and Postal Inspection Service reviews revealed a wide
array of efforts under way at post offices and districts.  Although
many of these efforts were innovative, postal headquarters had no
systematic way of sharing the results of successful efforts.  Two
mechanisms to facilitate information- sharing across the organization
had been developed but were not in use at the time of our review. 

  -- An Innovations Network, a computerized database, was set up to
     allow certain employee groups to share information on successful
     initiatives.  Coordinators at headquarters and in the field were
     to identify successful initiatives and submit descriptions of
     them for recording in the database. 

  -- A Customer Advisory Council Newsletter, published by the
     Consumer Affairs Department, was designed as a networking tool
     to be used by headquarters, field offices, and customer advisory
     councils that some post offices had established.  One purpose of
     the newsletter was to share the results of successful
     improvement efforts. 

According to postal headquarters officials, both of these mechanisms
were discontinued after the 1992 downsizing of the Postal Service. 
They said that after the downsizing, not enough employees were
available to maintain and promote these information-sharing efforts. 
In addition, the officials responsible for the Innovations Network
said that the procedures for maintaining and accessing the database
were cumbersome and that coordinators did not always update the
system to show new innovations.  We recognize that developing and
maintaining any system of sharing information on innovative
approaches to improving customer service will require resources.  The
cost of sharing such information would need to be weighed against the
benefits of giving all field offices the opportunity to implement
proven techniques for improving customer satisfaction. 

In our discussions with the former Vice President for Customer
Services, he said that a "clearinghouse" for new ideas and projects
was needed.  However, neither he nor other headquarters officials had
assigned responsibility for developing procedures to share
information on successful local initiatives. 


         RELATED INSPECTION
         SERVICE RECOMMENDATIONS
         NOT FULLY IMPLEMENTED
------------------------------------------------------ Chapter 3:5.3.1

In April 1993, the Postal Inspection Service reported a need for
better communication within and among district offices on successful
initiatives to improve customer services.  The report said that it
was not uncommon to find that some post offices had not shared
information on their improvement efforts with other post offices,
often in the same district.  The inspectors recommended in the April
1993 report that the Service take steps to permit sharing of such
information among post offices and districts.  They repeated the
recommendation in a December 1994 report, suggesting that postal
headquarters communicate CSI successes to offices nationwide via an
electronic message system. 

In responding to the latter report, the Vice President for Work Force
Planning and Service Management acknowledged that creating a bulletin
board for CSI users would potentially be useful.  Subsequently, his
office provided some information in the Service's automated
information system on best practices.  The system now identifies
those metropolitan areas with the highest average CSI ratings for
specific attributes, such as convenience of window service hours and
waiting time in line.  For example, 17 metropolitan areas were listed
for postal quarter 1, 1995, as having the highest average rating for
convenience of window service hours.  Users of the system are advised
that these areas are presumed to have put into place the best
practices for consistently meeting the needs of customers for this
service attribute.  The purpose of the information is to give those
interested in improving performance in particular attributes an idea
of where to go and whom to talk with about benchmark procedures
related to improvement efforts. 

Although it appears that this procedure for sharing information can
help, Service officials acknowledge that it falls short of fully
sharing information across the organization on practices found to
have worked best.  For example, the automated system does not
identify the practices followed by any of the metropolitan areas or
recognize the specific work teams responsible for new and innovative
practices that have proven successful. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 3:6

Although residential CSI results indicate that significant levels of
customer dissatisfaction continue to exist, the Postal Service is
taking the important first steps of adopting a policy of measuring
customer satisfaction to improve service.  Its numerous and promising
initiatives currently under way indicate a serious commitment to
overcoming policy, operational, and cultural barriers to improving
customer satisfaction by improving customer service. 

Although poor union-management relations constrain the Postal
Service, the development of a national strategy to focus all field
offices, including mail processing plants, on improving the
reliability of mail delivery service is a necessary step to
addressing a key cause of customer dissatisfaction.  Similarly, the
current performance incentive plans, which are innovative and a move
in the right direction, can be refined to give more emphasis to
encouraging prompt and timely mail delivery--what customers have said
that they want most from the Postal Service.  The Service could do
this by using measures of service reliability from EXFC and other
systems.  Because such data are already available, the added cost of
using these measures might be justified by potential benefits of
stronger focus by employees and management on improving service
reliability.  However, we recognize that the changes cannot be made
unilaterally for some employees.  For craft employees covered by SET,
changing the basis for the incentive payments would require agreement
with unions; for some other employees, the change would require
consultation with management associations. 

Many of the Service's national initiatives were relatively new, and
postal headquarters needs to know whether its initiatives are being
implemented and whether they are being done so in a timely manner. 
Without some system of tracking field offices' progress in
implementing such initiatives, headquarters officials cannot be sure
that field offices understand and are committed to the initiatives. 
Nor can officials systematically identify those offices most in need
of assistance and those adopting best practices and demonstrating
exceptional performance in implementing national initiatives.  The
Service would need to weigh the cost of implementing and maintaining
a system of sharing such information against the potential benefits
of improving customer satisfaction through better customer service. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 3:7

As part of the development of the Postal Service's national service
improvement strategy, and to achieve the greatest improvement in
customer satisfaction, we recommend that the Postmaster General take
the following steps: 

  -- Incorporate BCSI results in the Service's initiatives and
     ongoing efforts to improve its performance and service quality,
     using safeguards as appropriate. 

  -- Determine, in cooperation with unions and management
     associations, the feasibility of incorporating available
     measures of mail delivery service, along with CSI and other
     performance data, into employee pay incentive plans to encourage
     a stronger commitment to prompt and reliable mail delivery and,
     as appropriate, use these performance data in incentive plans. 

  -- Implement cost-effective procedures for headquarters units to
     use in monitoring and reporting the implementation and results
     of national service improvement initiatives to ensure that they
     are implemented as intended. 

  -- Implement cost-effective procedures for (1) regularly
     recognizing at the national level the best practices and
     successes of field offices and employees in improving customer
     satisfaction and (2) sharing information on such efforts across
     the organization. 


   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 3:8

In commenting on a draft of this report, the Service said that it
believed that our recommendations and concerns regarding employee
performance incentives, systematic implementation and monitoring of
improvements, and sharing best practices will be addressed in its
recently begun CustomerPerfect!  program.  To explain how that
program will address our recommendations, the Service discussed its
approach to identifying and sharing best practices.  It said that in
the past "best" had been a matter more of intuition than measurement,
and a team is looking at how to develop systems that will identify
possible best practices and validate their effectiveness by measuring
their results.  According to the Service, once a practice is
determined to be truly a best practice, it will be shared with the
field, possibly through electronic bulletin boards and presentations
at national or area-wide managers' meetings. 

The Service's CustomerPerfect!  initiative appears to be a reasonable
approach to addressing our findings and recommendations.  Moreover,
it is clear that the initiative has the commitment of the top-level
Postal Service leadership.  The program was just getting started at
the conclusion of our review, and it was too early to determine how
it will be implemented at lower management levels and by various
employee groups and how the program might affect delivery performance
and customer satisfaction. 

As noted in this report and our earlier report on labor-management
relations, the success of some earlier Service initiatives that were
designed to affect pay, duties, and management-employee relationships
of craft employees was limited by a lack of support from the unions
representing those employees.  At the time of our review, the Service
had not obtained the involvement and commitment of labor union
leaders in the CustomerPerfect!  initiative.  On the basis of the
Service's experience with similar past initiatives, we believe that
this involvement and commitment will be necessary to implement
aspects of the new initiative affecting craft employees and to
address our recommendation relating to the use of CSI and other
performance data, such as EXFC, in employee pay incentive plans. 




(See figure in printed edition.)Appendix I
CUSTOMER SATISFACTION INDEX SURVEY
QUESTIONNAIRE
============================================================ Chapter 3



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)


INFORMATION ON SIX CUSTOMER
SERVICE DISTRICTS VISITED BY GAO
========================================================== Appendix II

                     Central  Springfield          San
      Billings,      Plains,            ,   Francisco,     Chicago,    New York,
             MT           NB           MA           CA           IL           NY
--  -----------  -----------  -----------  -----------  -----------  -----------
Po          331        1,112          415          239           53          117
 st
 o
 f
 f
 i
 c
 e
 s
 ,
 s
 t
 a
 t
 i
 o
 n
 s
 ,
 a
 n
 d
 b
 r
 a
 n
 c
 h
 e
 s
Em        2,300        9,100        6,900        7,400        6,300       10,800
 p
 l
 o
 y
 e
 e
 s
 \
 a
Po      152,000      341,000      180,000      218,000       72,000      123,000
 st
 o
 f
 f
 i
 c
 e
 b
 o
 x
 e
 s
 \
 b
De
 l
 i
 v
 e
 r
 i
 e
 s
 :
 \
 b
Ci      202,000      976,000      354,000    1,079,000    1,166,000    1,251,000
 ty
Ru       90,000      371,000      180,000       65,000            0            0
 r
 a
 l
 \
 c
To      292,000    1,347,000      534,000    1,144,000    1,166,000    1,251,000
 t
 a
 l
--------------------------------------------------------------------------------
\a Rounded to nearest hundred. 

\b Rounded to nearest thousand. 

\c Includes highway contract route deliveries. 

Source:  Compiled by GAO from U.S.  Postal Service data. 


SURVEY RESULTS
========================================================= Appendix III



                              Table III.1
                
                 CSI Results for Metropolitan Areas in
                 Postal Quarters 1, 2, and 3 of Fiscal
                 Year 1994 Include a Range of Sampling
                                 Errors


CSI question                                        Smallest   Largest
--------------------------------------------------  --------  --------
Customer's post office
----------------------------------------------------------------------
Courtesy of window clerks                                1.6       3.6
Helpfulness of window clerks                             1.5       3.5
Waiting time in line                                     2.3       3.9
Convenience of window service hours                      2.2       3.7
Availability of stamps through                           2.8       4.6
 vending machines
Having vending machines in working                       2.7       4.9
 order
Convenience of lobby hours                               1.7       4.8
General inside appearance of building                    1.2       3.5
Availability of parking at or near post office           2.2       3.9

Mail delivery to customer post office box
----------------------------------------------------------------------
Delivery of mail to box by                               2.8      16.4
 scheduled time
Delivery of mail to correct box                          3.0      17.3

Forwarding/change of address service
----------------------------------------------------------------------
Forwarding mail within a reasonable number of days       6.3      11.5
Forwarding mail to correct person                        5.7      11.1
Prompt start-up of delivery to new                       5.7      10.7
 address

Telephone experience
----------------------------------------------------------------------
Ease of getting through                                  3.7       6.7
Speed of answering phone                                 3.6       6.6
Ability of person who answered to help or refer          3.6       6.4
 call
Courtesy of employees                                    3.2       5.8
Accuracy of information given                            3.2       6.1

Complaint handling
----------------------------------------------------------------------
Making it easy to complain or describe your              5.4      12.2
 problem
Speed of response to your problem                        4.4      12.4
How well you were dealt with                             4.9      12.7
----------------------------------------------------------------------
Source:  Compiled by GAO from U.S.  Postal Service data. 



                              Table III.2
                
                Selected Survey Results for Metropolitan
                Areas With Sampling Errors in Quarter 3,
                                  1994


                                                    At least  At least
Question on Postal Service performance in the past     ï¿½ 7.6    ï¿½ 10.0
3 months\a                                           percent   percent
--------------------------------------------------  --------  --------
Mail delivery to customer's post office box
----------------------------------------------------------------------
Delivery of mail to box by scheduled time                 36        12
Delivery of mail to correct box                           49        14

Forwarding/change of address service
----------------------------------------------------------------------
Forwarding mail within a reasonable number of days       155        14
Forwarding mail to correct person                        133         7
Prompt start-up of delivery to new                       114         5
 address

Complaint handling
----------------------------------------------------------------------
Making it easy to complain or describe                   159        53
 your problem
Speed of response to your problem                        155        68
How well you were dealt with                             158        77
----------------------------------------------------------------------
\a Other questions are not shown because no metropolitan area had
sampling errors of at least 7.6 percentage points in the third
quarter of 1994. 

Source:  Compiled by GAO from U.S.  Postal Service data. 


INFORMATION ON POSTAL SERVICE
PERFORMANCE INCENTIVE PAYMENTS IN
1994 AND 1995
========================================================== Appendix IV

The Postal Service makes team-based performance incentive payments to
certain craft employees, Executive and Administrative Schedule (EAS)
employees, and most Postal Career Executive Service (PCES).  The
Service bases the payments to all three categories of employees in
part on the results of quarter 4 CSI results, using only excellent
and very good responses. 


   CRAFT EMPLOYEES
-------------------------------------------------------- Appendix IV:1

The Postal Service and unions representing rural carriers, mail
handlers, and postal police officers agreed to an incentive plan,
called Striving for Excellence Together (SET), in 1991.  The first
payments under SET were made in January 1993 on the basis of fiscal
year 1992 performance.  For the most recent SET payments based on
fiscal year 1994 performance, 92,854 rural carrier employees received
incentive payments under SET.  These employees account for about 13
percent of the Service's approximately 700,000 craft employees. 
Clerks, city carriers, and mail handlers account for most of those
employees not participating.\1 The unions representing clerk and city
carrier employees rejected the Service's proposal during 1990-1991
contract negotiations to join SET because they believed that such pay
would replace negotiated wage increases and also encourage
competition among employees.\2

SET payments are in addition to regular wages, cost-of-living
adjustments, and overtime pay available to craft employees. 
Employees receive varying SET payments each January on the basis of
three factors:  improvement in the Service's national financial
performance compared to the prior year, as measured by increases in
the ratio of total revenue to total paid hours; the performance
cluster's ranking in the overall CSI rating compared to other
clusters; and the extent of the cluster's improvement in CSI rating
compared to the previous year. 

We obtained information on SET payments in January 1994 and January
1995 for the two most recent fiscal years, 1993 and 1994.  Payments
under SET were higher in January 1994 than January 1995 primarily
because of a general decline in CSI ratings between the quarter 4
ratings.  The average SET payment was $509 to participating employees
in January 1994 and $82.04 in January 1995 based on fiscal years 1993
and 1994 results.  The payment to each SET-covered employee for the
Service's national financial performance was $300 in January 1994. 
In January 1995, craft employees received no incentive payment based
on financial performance.  The total SET payments in January 1994
ranged from a low of $335 for the Chicago, IL, and Philadelphia, PA,
performance clusters to a high of $730 for the Providence, RI,
performance cluster.  In January 1995, the total payments ranged from
$0 for nine performance clusters to $365 for the Middlesex-Central,
MA, cluster. 


--------------------
\1 In fiscal year 1993, additional craft employees, namely mail
handlers and postal police, participated in the SET program along
with rural carriers. 

\2 Our report entitled U.S.  Postal Service:  Labor-Management
Relations Problems Persist on the Workroom Floor (GAO/GGD-94-201 A
and B, Sept.  29, 1994) provides additional information on the lack
of union participation in SET and various other Postal Service
improvement initiatives. 


   EAS EMPLOYEES
-------------------------------------------------------- Appendix IV:2

In 1993, after consultation with management associations representing
some of the affected employees, the Service began replacing its merit
pay plan for EAS employees with a new plan linking merit pay
increases to the SET payments discussed above.\\3 EAS employees
include postmasters, supervisors, certain executives, and
administrative personnel.  Some EAS employees transitioned to the new
plan beginning with payments in January 1994 on the basis of fiscal
year 1993 performance, and all EAS employees were covered the
following year.  The payments under the plan include a one-time SET
payment in January each year and a merit increase as a percentage of
each employee's base pay, both of which are determined on the basis
of the three SET factors mentioned earlier. 

As a result of fiscal year 1993 ratings, in January 1994, about
13,500 EAS employees covered by the plan received a one-time SET
payment, which averaged $509 nationally and followed the payment
ranges by cluster of the SET payments to certain craft employees. 
The permanent pay increase ranged from 3.9 percent--the minimum for
any location--up to 5.6 percent for covered employees in Providence,
RI.  The January 1995 SET payments were made to about 77,000 EAS
employees, averaged $82.02, and ranged from $0 for 9 clusters up to
$365 in 2 clusters (Middlesex-Central, MA, and New Hampshire).  The
permanent pay adjustment ranged from the minimum of 0.6 percent for
49 clusters to a high of 2.8 percent for 1 cluster
(Middlesex-Central, MA). 


--------------------
\3 The new merit pay plan also has a cash award provision for
recognizing exceptional individual performance. 


   PCES
-------------------------------------------------------- Appendix IV:3

The Postal Service began making performance incentive payments to
career executives on the basis of CSI results and other measures in
January 1993, based on fiscal year 1992 performance.  PCES-I
executives are covered by the incentive plan.  These executives,
totaling 956 in 1994, generally are managers responsible for national
policies and managers responsible for major organizational units. 
PCES-II executives, totaling 34 in 1994, who are the corporate
officers and responsible for broad functional departments and the 10
postal area offices, are not covered by the plan.  The incentive
payments are based on local and national measures of financial
performance, i.e., budgeted versus actual net income or deficit for
the year as well as the results of CSI and EOS surveys. 

In January 1994, all PCES-I executives received SET payments
averaging $509 for fiscal year 1993 performance and a minimum 3.9
percent merit payment, the same as EAS employees for that year.  For
fiscal year 1994 performance, the Postal Service incorporated several
new design principles into the incentive payment plan for PCES-I
executives.  These principles included a measure of the "voice of the
employee" as indicated through annual EOS results, factors relating
to continuous improvement, and exclusion of payments to executives in
areas with poor performance.  Of 954 PCES-I executives, 922 received
incentive awards in January 1995.  The awards averaged 2.4 percent of
base pay nationally and ranged from no increase for 32 executives up
to 12.1 percent for 10 executives. 

According to Postal Service officials, incentive payments to PCES-II
corporate officers did not consider the same financial, CSI, and EOS
results as used for PCES-I executives through fiscal year 1994. 
However, they said that the Board of Governors approved an incentive
plan for PCES-II executives for fiscal year 1995 performance. 


SELECTED SERVICE IMPROVEMENT
INITIATIVES IN SIX CUSTOMER
SERVICE DISTRICTS VISITED BY GAO
IN 1994
=========================================================== Appendix V

Service factor and
initiative          Description
------------------  ------------------------------------------------------------
Window service
--------------------------------------------------------------------------------
"Tax Night" hours   Maintain extra staff and extended hours on April 15 to
                    assist customers in mailing their tax returns.

Sunday hours        Provide window service on Sundays for the December holiday
during holiday      season.
season

Package pick-up     Allow customers to pick up packages and certified mail on
when post office    Saturdays even though window service is not provided.
is closed

Lobby directors     Use an employee in the lobby to guide customers to the
                    proper window service line or proper forms and sell stamp
                    booklets.

Special lines for   Maintain separate window service lines for transactions that
long transactions   take a significant amount of time (i.e., setting postal
                    meters, passport applications, and money order sales).

Appointments for    City post office reduces waiting time for postage meter
setting meters      customers and others by allowing the meter customer to come
                    and have the post office reset the postage meter at an
                    agreed time.

Disabled sent to    Allow disabled customers to come to the front of the window
front of line       service line.

Setting meters at   Postal employee visits postage meter customer's place of
place of business   business to reset meter.

Supervisors         On the basis of a complaint from APWU, window service
trained on          supervisors at one post office attended training to learn
integrated retail   how to use the clerks' retail terminals so they could
terminals           understand and help resolve clerk and customer problems.

Meters placed in    City post office allows postage meter customers to drop off
parcel lockers      their meters and pick up the reset ones from post office
when set            parcel lockers at their convenience.

Supervisors         City post office allows supervisors to reset postage meters
allowed to set      when window service lines become long.
meters

Supervisor          Post office supervisors monitor length of window service
monitors lines      lines by looking through a window facing the window lines or
through window or   a video camera.
video

Drive-through       Suburban post office allows customer to drive his or her car
windows             to the side of the post office to transact business (similar
                    to a bank drive-through window).

Resource or unit    Post office operations managers conduct reviews of their
reviews for         post offices' staffing and scheduling to help postmasters
adequate staffing   put employees in the right assignments at the right time to
                    serve customers.

Select additional   City post offices get authority to select additional window
window clerks       clerks to replace those who retired during the
                    reorganization.

More clerks at      During the busy times of the day, post office staffs
busy times          additional windows to serve customers to reduce waiting
                    time.

Redeploy vending    District office allows postage vending machines to be placed
machines off-site   at off-site locations instead of being retired. This
                    increases customer convenience and keeps post office window
                    service lines shorter.

Centralized         District provides a central location for district customers
stamps-by-mail      to call or to write for purchasing postage stamps.
unit

Parcel lockers in   District places parcel mail lockers in high-rise apartment
apartment           buildings so that if delivery to a customer cannot be made,
buildings           the package is left in the locker for pick-up at customer's
                    convenience.

Mystery shopping    District staff conducts surprise visits and monitors post
                    office window service lines to determine if window clerks
                    provide good customer service.


Lobby service
--------------------------------------------------------------------------------
Time locks on       Vending machine and post office box lobby door
lobby doors         electronically locks at a set time even though an employee
                    is not present.

Card access to      Post offices provide customers with plastic cards that allow
lobbies             them to enter the post office lobby to use postal vending
                    machines and have access to their post office boxes when
                    post office is closed.

Vending machine     Placement of a postal vending machine outside of a suburban
placed outside of   post office to allow customers to purchase postal products
building            24 hours a day and 7 days a week.

Some lobbies open   Lobby doors kept unlocked 24 hours a day for customers to
24 hours            access their post office boxes and vending machines.

Self-service        City post office maintains a self-service weighing scale in
scales              the lobby for customers' convenience. This allows the
                    customers to determine the correct postage before using a
                    vending machine or window service.

Hire more           Hire additional postage vending machine technicians to
technicians         service the machines.

Mystery shopping/   District monitors post office lobby service by surprise
lobby inspections   inspection visits.



Telephone experience
--------------------------------------------------------------------------------
Postal service      District makes telephone courtesy and helpfulness training
telephone training  available to all local post offices on video tape.

Telephone company-  Telephone equipment company provides training on courtesy
provided training   and helpfulness when a new system is installed in an office.

Rollover and hold   Office telephone systems are upgraded so customer calls
features installed  "roll over" to a free telephone line that allows employees
                    to place customers on hold. Customers hear information about
                    Postal Service products while on hold.

Take advantage of   District takes advantage of the Postal Service's national
national service    telephone equipment service contracts so that small post
contracts           offices receive prompt repair service at the best possible
                    price.

Install answering   Small rural post office uses telephone answering machines to
machines in         provide information to customers when postmaster is away for
offices that close  lunch.
for lunch

Voice mail in       Allows customer to leave messages so that customer does not
larger offices      have to take time to call again.

"No transfer"       District policy prohibits postal clerks from referring
policy              customer calls for information to another postal employee.

Centralized         All calls to post offices or stations in a selected
telephone inquiry   geographic area are answered at a central location where
number              clerks have on-line computer access to ZIP Code and other
                    Postal Service information.

"Answer book"       District office or work team summarizes answers to
                    frequently asked customer questions and distributes to all
                    post offices for reference.

Computerized ZIP    Personal computer disk that contains all ZIP Codes with
code directory      corresponding locations.

Computerized        Personal computer disk that contains the domestic mail
domestic mail       manual is formatted for easy access.
manual

Postal answer line  Postal answer line tape is replaced with modern "touch tone"
                    accessed system, or updated to reflect recent organizational
                    changes.

Put telephone near  Window clerk does not need to leave his or her workstation
integrated retail   to answer the telephone.
terminals

Survey post         District makes effort to determine telephone service
offices to          availability and needs at its post offices.
determine
telephone needs

Designate           District Management By Participation team designates
telephone service   district personnel who can expertly handle postmaster
"experts"           questions in a particular area.

Three-ring policy   District and post office policy is for staff to answer
                    customer telephone calls within three rings.

Mystery calling     District staff makes a surprise call to a post office to
                    determine if the telephone clerk provides good customer
                    service.


Complaint handling service
--------------------------------------------------------------------------------
"45-day" automated  District Consumer Affairs Office's computer system
follow-up letter    automatically prints a letter to the customer asking whether
                    his or her complaint has been satisfactorily resolved.

Personal contact    Postmasters contact customers with complaints to show
by postmaster       personal interest in resolving them.

Resolve complaints  District policy for post offices is to resolve complaints at
at lowest level     post offices rather than at the district level.

"Call back by 2"    City post office has a pilot program in which complainants
program             receive a personal telephone call by 2 PM on the day that
                    the complaint is received. The postmaster sends a letter the
                    same day if the customer cannot be reached by telephone.

On-line complaint   District Consumer Affairs Office maintains automated
data system         complaint data system to readily inform an inquiring
                    customer of the status of his/her complaint.


Post office property
--------------------------------------------------------------------------------
"Modest"            Post offices make a range of inexpensive improvements to
renovations to      their inside appearance, utilizing equipment and materials
standardize         available in the post office, or nearby post offices. The
lobbies             renovations may include painting and standardized signs.

Retail specialist   Retail specialists at the district office approve all new or
approval of all     renovated retail space.
renovated retail
space

Move post offices   Postmasters request emergency exceptions to the USPS capital
out of substandard  construction freeze to move, or extensively renovate, post
buildings           offices with health and safety violations.

Lobby inspections   District monitoring effort to improve lobby appearance.

Postmaster self-    District postmasters self-evaluate their customer services
evaluations         through a checklist.

MBP "expert"        District Management By Participation team designates
assistance          district personnel who can expertly handle postmaster
                    questions in a particular area.

Postmaster credit   Districts will formalize use of credit cards by postmasters
cards for cleaning  for small purchases to bypass an extended procurement
and related         process.
supplies

"15-minute"         Customer parking at a post office or station is limited to
parking             15 minutes to alleviate congestion.

Off-site parking    The postmaster leases off-site space where his or her
for employees,      employees are to park so customers can park at the post
leaving on-site     office.
parking for
customers

Left hand drops to  Post offices have mail collection boxes accessible to
reduce need for     driver's side window of a car so that the driver can deposit
parking             mail without parking or entering the post office.

Hiring additional   District hires custodial workers to maintain clean post
custodians          offices.


Post office box service
--------------------------------------------------------------------------------
Increased           Post office supervisors look at post office box mail to
supervision to      determine if it was accurately filed.
detect missorts

Have plant sort     Plant sorting of mail in an effective manner helps the local
box mail by         post office fill the post office boxes by a set time.
sections to help
meet "up time"

Supervisors ensure  Supervisors monitor the amount of mail that needs to be put
there are           in post office boxes by a set time and obtain other post
sufficient clerks   office clerks to complete the process if necessary.
to meet "up time"

Resource or unit    Post office operations managers conduct reviews of their
reviews to ensure   post offices' staffing and scheduling to help postmasters
that clerks will    ensure that clerks are available to put mail in the post
be available to     office boxes on time.
put up mail

Parcel lockers for  Post office installs parcel mail lockers with regular post
box section         office boxes for customer convenience (no need to wait on
                    window line).

Install additional  Provide more post office boxes during a lobby renovation.
box section

6,000-box post      A Postal Service station provides post office box and
office box station  general delivery service in a community with a large
in area with large  homeless population.
homeless
population

"Box Call" program  Employees electronically record which post office boxes have
                    mail so customers can use a touch tone telephone to find out
                    whether or not they need to pick up their mail.

Wooden dowel or     Clerks mark a box section with a wooden dowel, or a light,
light marks         when they have finished putting mail in the boxes so
sections that are   customers know that their mail is available.
"up"

Place boxes         Some or all post office boxes are physically accessible from
outside the         outside the building so customers can pick up mail 24 hours
building            per day.

"Box Activity       Personal computer program that records and tracks post
Tracking System"    office box availability, rent due date, and customer names.
                    The system also prints rent due notices for customers.

Area-wide "up       Area office establishes a standard "up time" for all post
time"               office box service.


Forwarding/
change of address service
--------------------------------------------------------------------------------
Carrier manually    Letter carrier forwards customer mail to a new address if
forwards until CFS  computer system does not list customer's new address for
"catches up"        automatic mail forwarding.

Double-check mail   Post office supervisor or postmaster checks forwarding mail
before sending to   for accuracy before sending on to district's automated mail
forwarding unit     forwarding unit.

Change of address   District provides customer with new address labels in his
package with        postal change of address package.
labels

Employee receiving  Employee reviews change of address cards for completeness
change of address   and legibility before accepting them from customers.
card from
customers
responsible for
quality control

No-record mail      Team of employees and managers takes responsibility for
task teams          reducing no-record mail volume by using quality improvement
                    techniques, including visits to the mail forwarding sites
                    and double-checking all mail to be sent to the sites.

Area CFS focus      Area office sets up a group of managers from different
group               district offices to brainstorm possible solutions to CFS
                    problems.

Street address      District effort to aid the city government in ensuring that
project with city   it has the correct address for city residents.

Verify new          Small rural post office ensures that public utilities have
addresses with      the correct addresses for post office customers.
utilities

Improve working     Provide the mail forwarding site with adequate work space
conditions at       and air conditioning so employees can work effectively.
forwarding site

CFS training for    Train letter carriers on proper change of address and mail
carriers            forwarding procedures.

CFS training/       District arranges for postmasters with good no-record mail
assistance for      rates to provide guidance or training to those with higher
postmasters         rates.

CFS "rolodex" for   All CFS cards are filed in a centralized rolodex for letter
carriers            carriers.

File CFS cards      Clerks place CFS cards directly in letter carriers' cases
directly in         instead of in a central file.
carrier cases


Miscellaneous
(indirect CSI improvements)
--------------------------------------------------------------------------------
Local customer      District obtains customer feedback at the post office level.
surveys providing
5-digit data,
(including
opinionmeter,
lobby surveys,
mail-out surveys)

Have postmasters/   Postmasters and station managers are in the lobby every day
station             for a few hours to provide a better image to customers and
managers in the     provide personal feedback for their concerns.
lobby

CSI commitment      District staff visits post office and monitors customer
worksheets          service.

Clerks speak a      City post office with a large number of immigrant customers
second language     has a clerk available to translate for them.

Managers            District post office managers participate in community
participate in      affairs so that community perception of the Postal Service
civic activities    may be higher.

"Town Meetings"     A postmaster plans to hold meetings for all customers--not
with customers      just those who complain--at a different station each month
                    to listen to their concerns.

"Pride in           Postmaster presents awards to letter carriers who have no
Delivery"           misdeliveries in a 6-month period. Customers call a special
misdelivery         telephone number for reporting misdelivered mail, which a
program             supervisor picks up and redelivers.

Hire postmaster     Post offices that are ordinarily staffed by one person--the
reliefs for small   postmaster--are allowed to hire a relief postmaster at an
offices             hourly rate to serve customers when the full-time postmaster
                    is out of the office.

Focus groups        Postmaster or other manager convenes a group of customers to
                    discuss a specific issue.

CSI video           A 12-minute videotape prepared by an area office is designed
                    to explain CSI to craft employees who may not be familiar
                    with it and who may not have direct customer contact.

Educational         Postmasters send letters to their customers describing the
letters to          steps that they are taking to improve service.
customers
--------------------------------------------------------------------------------



(See figure in printed edition.)Appendix VI
COMMENTS FROM THE U.S.  POSTAL
SERVICE
=========================================================== Appendix V



(See figure in printed edition.)


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix VII


   GENERAL GOVERNMENT DIVISION,
   WASHINGTON, D.C. 
------------------------------------------------------- Appendix VII:1

Michael E.  Motley
James T.  Campbell
Kenneth E.  John
Lawrence R.  Keller


   NEW YORK REGIONAL OFFICE
------------------------------------------------------- Appendix VII:2

James D.  VanBlarcom
Frank Grossman
John Schick

*** End of document. ***