Postal Service: Service and Cost Aspects of the Dakota Central Area Mail
Processing Center (Letter Report, 10/18/94, GAO/GGD-95-16).

This report reviews mail service in the communities served by the Dakota
Central Area Mail Processing Center in Huron, South Dakota. GAO assesses
the validity of concerns of postal customers that the new center, which
opened in July 1993, would not be able to maintain overnight mail
service in the area. GAO compares mail service in Aberdeen, South
Dakota, which processes its own mail, with service in and between other
communities whose mail is processed at the center. GAO also compares (1)
the cost of the center with the projected costs and (2) the number of
personnel and workhours devoted to mail processing before and after the
consolidation.

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

 REPORTNUM:  GGD-95-16
     TITLE:  Postal Service: Service and Cost Aspects of the Dakota 
             Central Area Mail Processing Center
      DATE:  10/18/94
   SUBJECT:  Postal service
             Mail transportation operations
             Mail delivery problems
             Work measurement standards
             Cost control
             Construction costs
             Administrative costs
             Postal facilities
             Personnel management
             Productivity
IDENTIFIER:  Huron (SD)
             Aberdeen (SD)
             Watertown (SD)
             Mitchell (SD)
             
**************************************************************************
* 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.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


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


Report to the Chairman, Committee on Appropriations, U.S.  Senate

October 1994

POSTAL SERVICE - SERVICE AND COST
ASPECTS OF THE DAKOTA CENTRAL AREA
MAIL PROCESSING CENTER

GAO/GGD-95-16

Dakota Central Center


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


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


B-251104.2

October 18, 1994

The Honorable Robert C.  Byrd
Chairman, Committee on Appropriations
United States Senate

Dear Mr.  Chairman: 

This report responds to the Committee's directive in Senate Report
102-353 that we review mail service in the communities served by the
Dakota Central Area Mail Processing Center in Huron, SD.  We assessed
the validity of the concerns of postal customers that the new Center,
which opened in July 1993, would not be able to maintain overnight
mail service in the area.  We tested and compared mail service in
Aberdeen, SD, which processes its own mail, with service in and
between other communities whose mail is processed at the Center.  In
addition, because the main purpose of the Center was to save money,
we compared (1) the cost of the Center with the projected costs and
(2) the number of personnel and workhours devoted to mail processing
before and after the consolidation. 


   BACKGROUND
------------------------------------------------------------ Letter :1

This is our second report on the Dakota Central Center.  The first,\1
issued before the Center opened, showed that (1) the consolidation
would slow mail service to certain locations--this was not disclosed
in the Postal Service's consolidation plan for the Center--and (2)
the plan overstated improvements in mail service that were to occur. 
Because of the concerns about deterioration in service, the Postal
Service later modified the consolidation by dropping Aberdeen, SD,
from the communities whose mail would be trucked to the Center for
processing.\2 As a result, the principal post offices in the
consolidations are Huron, Watertown, and Mitchell, SD.  Figure I.1
shows the overnight service area for the Dakota Central Center and
the communities where we tested mail service. 

The benefits traditionally expected of consolidated area mail
processing centers, as outlined in postal guidelines for their
development and implementation, have been (1) a savings in mail
processing costs and (2) the maintenance of the same or better
delivery service.  In October 1992, postal headquarters suspended
action on new consolidation proposals to examine the concept and
revise guidelines to ensure that they supported the Service's
priorities for "customer service, employee obligations, and operating
efficiency." At the conclusion of our work, the Service was revising
and testing new guidelines for development, implementation, and
evaluation of the performance of area mail processing centers. 


--------------------
\1 See Postal Service:  Service Impact of South Dakota Mail Facility
Not Fully Recognized (GAO/GGD-93-62, Feb.  25, 1993). 

\2 This modification eliminated some of the situations where service
would be slowed; however, Watertown lost overnight service to Fargo,
ND, and Wilmar, MN, because of the consolidation. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :2

Our mail test of First Class letters showed that Aberdeen, which
processes its own mail, had slightly better overnight delivery than
the communities served by the Dakota Central Center.  However,
service in and between the communities served by the Center was
generally within the range of the Postal Service's performance
standard for overnight delivery and substantially exceeded most other
performance elsewhere in the nation during the period we did our
test. 

Construction costs of the Center as well as first-year operating
costs significantly exceeded estimates in the consolidation plan. 
The Service had planned to renovate an existing building for the
Center at a cost of about $624,000 but instead designed and built a
new facility on newly purchased land at a cost of about $3 million. 
Although the Service projected that the Center would reduce personnel
costs, the number of employees and workhours devoted to mail
processing and customer service increased by 44 percent compared to
the year before, when mail processing operations were decentralized. 

Because of the ongoing concern over service among postal customers in
this area, the cost difficulties encountered at the Center, and the
possibility that the Service may bring more communities under it, the
Postmaster General agreed with our recommendation to do a
postimplementation review of the Dakota Central Center for both
delivery performance and cost. 


   OBJECTIVE, SCOPE, AND
   METHODOLOGY
------------------------------------------------------------ Letter :3

Our objective was to determine if overnight delivery service was
being provided in the area served by the Dakota Central Area Mail
Processing Center.  We also compared certain projected and actual
costs related to establishing operations at the Center. 

To assess overnight First Class mail delivery service, we mailed
1,321 First Class letters from a judgmental sample of locations
within the 1-day service area of Dakota Central.  We selected for our
test the three principal communities in the consolidated area; Huron,
Mitchell, and Watertown.  To test for differences in service between
the consolidated communities and one that was not included in the
plan, we included Aberdeen.  We arbitrarily selected 12 outlying
towns and sent 3 letters to each of the 4 principal communities for a
total of 144 letters.  We also selected arbitrarily 8 remote towns
that were included in the consolidation and sent a total of 48 First
Class letters between these towns.  Mailings were sent from a
judgmental sample of mail boxes and post offices in each community to
a single address in the recipient community.  Table 1 shows our test
design. 



                           Table 1
           
           Communities in the Mail Test and Number
                      of Letters Mailed


                                   Mitchel  Watertow  Remote
From              Aberdeen  Huron        l         n    town
----------------  --------  -----  -------  --------  ------
Aberdeen               270    N/A      N/A       N/A     N/A
Huron                 15\a    270      N/A       N/A     N/A
Mitchell              18\a    9\a       61       N/A     N/A
Watertown             21\a   12\a     14\a     405\b     N/A
Outlying town         36\a   36\a     36\a      36\a     N/A
Remote town            N/A    N/A      N/A       N/A    48\a
------------------------------------------------------------
Legend:  N/A = There was no mailing to this community from the other. 

\a Indicates a lack of generalizability due to number of mailings
and/or the method used to select the community.  Results only
demonstrate a possible trend, not conclusive evidence. 

\b Mailings took place on two occasions, one in February 1994 and one
in March 1994, because of an addressing error. 

To determine rates for overnight delivery, we determined the
percentage of the test mailings that arrived overnight and compared
this performance to the Service's goal of 95 percent as well as to
the average on-time performance nationwide for 1-day mail during the
same period.  See appendix I for a more detailed description of the
mail test. 

To compare projected and actual costs of the new facility, we
analyzed the Dakota Central consolidation plan and various financial
and service records on the construction and operation of the new
facility.  We discussed the implementation of the Center and its
effect on costs and service with postal officials in communities
affected by the consolidation.  We also reviewed workhour and other
data on postal operations in Huron, Mitchell, and Watertown, SD,
before and after the Center opened. 

One purpose of installing automation in a center such as Dakota
Central is to allow barcodes to be applied to letters, which should
reduce sorting costs at other automated processing centers where the
mail is destined.  We determined how certain processing costs were
affected in the Dakota Central area but did not determine how Dakota
Central barcoding affected processing costs at locations outside the
Dakota Central area.  Similarly, we did not determine delivery times
for Dakota Central mail destined outside of the overnight delivery
area. 

We did our review from December 1993 through June 1994 in accordance
with generally accepted government auditing standards.  The
Postmaster General provided written comments on a draft of this
report, which are summarized at the end of this letter and reprinted
in appendix II. 


   ON-TIME SERVICE GENERALLY
   ATTAINED IN THE OVERNIGHT
   DELIVERY AREA
------------------------------------------------------------ Letter :4

We tested actual delivery times for First Class letters mailed within
and between the four principal communities of Aberdeen, Huron,
Watertown, and Mitchell, SD.  Aberdeen provided a benchmark because
it was not included in the consolidation, and its mail was processed
at the local post office.  Mail from the other 3 communities went to
the Dakota Central Center in Huron for processing, a round-trip
distance of 184 miles for Watertown, 104 miles for Mitchell, and 4
miles for Huron. 

Aberdeen had the best performance, with over 99 percent of our
letters received the next day.  Watertown had an overall overnight
delivery rate of 97 percent, and Huron had 93 percent.  Our results
in Mitchell were much lower; only about 57 percent of our letters
were received the next day.  Initially, postal officials were unable
to explain why the performance in Mitchell was so low.  However,
after some research, they said that the delayed letters, which
carried a street address, were delivered to a post office box
maintained by the addressee and that the delay occurred between the
post office box and our receipt at the street address.  They believed
that the letters were likely available at the addressee's post office
box the day after being mailed and that Mitchell's overnight service
was higher than our test indicated. 

Ninety-two percent of letters mailed between communities arrived
overnight.  However, because of the uncertainty regarding the
Mitchell test results, the actual rate may be higher than 92 percent. 
The specific results are shown in table 2. 



                                     Table 2
                     
                      Results of Mailings Within and Between
                            the Principal Communities



                    Number            Number            Number            Number
                  received          received          received          received
          Number  overnigh  Number  overnigh  Number  overnigh  Number  overnigh
From      mailed         t  mailed         t  mailed         t  mailed         t
--------  ------  --------  ------  --------  ------  --------  ------  --------
Aberdeen     270       269     N/A       N/A     N/A       N/A     N/A       N/A
Huron         15        15     270       252      10         6      12        12
Mitchell      18        18       9         8      61        35      12        12
Watertow      21        21      12        12      14         9
 n                                                                 270       266
 First                                                             135       125
 mailing
 Second
 mailing
--------------------------------------------------------------------------------
Legend:  N/A = There was no mailing to this community from the other. 

Overall, these results compare favorably to the Service's goal of 95
percent\3 for on-time delivery and the average of 79 percent for
overnight mail in urban areas throughout the nation during that
quarter.\4

We also tested the mail delivery from and between smaller towns
adjacent to these communities, including those near the end of the
principal mail routes in the area.  Although the number of mailings
included in this part of the test was too small to provide conclusive
information, we did these mailings for general insight into whether
there might be delivery problems in these areas.  These results are
shown in tables 3 and 4. 



                                     Table 3
                     
                     Results of Mailings From Outlying Towns
                             to Principal Communities



                    Number            Number            Number            Number
                  received          received          received          received
          Number  overnigh  Number  overnigh  Number  overnigh  Number  overnigh
From      mailed         t  mailed         t  mailed         t  mailed         t
--------  ------  --------  ------  --------  ------  --------  ------  --------
Groton         3         3       3         3       3         0       3         0
Langford       3         3       3         0       3         0       3         0
Britton        3         3       3         3       3         0       3         0
Wessingt       3         2       3         3       3         2       3         3
 on
Wolsey         3         2       3         3       3         2       3         2
Hitchcoc       3         2       3         2       3         2       3         3
 k
Alexandr       3         3       3         3       3         3       3         3
 ia
Bridgewa       3         3       3         3       3         3       3         3
 ter
Parkston       3         3       3         2       3         3       3         3
De Smet        3         0       3         0       3         0       3         0
Henry          3         3       3         3       3         3       3         0
Willow         3         3       3         3       3         3       3         0
 Lake
--------------------------------------------------------------------------------


                                                                       Table 4
                                                       
                                                       Results of Mailings Between Remote Towns



                              Number                Number                  Number                 Number                Number                Number
                  Number    received    Number    received      Number    received     Number    received    Number    received    Number    received
From              mailed   overnight    mailed   overnight      mailed   overnight     mailed   overnight    mailed   overnight    mailed   overnight
--------------  --------  ----------  --------  ----------  ----------  ----------  ---------  ----------  --------  ----------  --------  ----------
Big Stone City         3           3         3           3         N/A         N/A        N/A         N/A       N/A         N/A       N/A         N/A
Eureka               N/A         N/A       N/A         N/A           3           0          3           3       N/A         N/A       N/A         N/A
Fort Thompson        N/A         N/A       N/A         N/A         N/A         N/A        N/A         N/A         3           3         3           3
Conde                  3           2         3           3         N/A         N/A        N/A         N/A       N/A         N/A       N/A         N/A
Reliance             N/A         N/A       N/A         N/A         N/A         N/A        N/A         N/A         3           3         3           2
Hurley                 3           3         3           3         N/A         N/A        N/A         N/A       N/A         N/A       N/A         N/A
Gary                   3           3         3           3         N/A         N/A        N/A         N/A       N/A         N/A       N/A         N/A
Rosholt              N/A         N/A       N/A         N/A           3           0          3           3       N/A         N/A       N/A         N/A
-----------------------------------------------------------------------------------------------------------------------------------------------------
Legend:  N/A = There was no mailing to this community from the other. 


--------------------
\3 The 95-percent confidence intervals applicable to our tests are
shown on page 15. 

\4 The national average is based on the quarterly measurement of
First Class mail service by Price Waterhouse under contract to the
Postal Service.  Delivery time is tracked in 96 cities nationwide. 
During the period of our review, none of the cities met the
95-percent standard; performance ranged from 52 percent in New York
City to 92 percent in Wichita. 


   START-UP AND OPERATING COSTS
   HAVE SIGNIFICANTLY EXCEEDED
   ESTIMATES
------------------------------------------------------------ Letter :5

Estimated first-year start-up costs for the Center in the
consolidation plan were $1.46 million.  Actual costs were $3.26
million.  The principal reason for the large difference is that the
plan called for renovation of an existing building in Huron to house
the Center.  However, after the plan was approved by postal
headquarters, the Service decided to purchase land and construct a
new building, at a cost of about $3 million versus the $624,000
planned for renovation.  A district official in Sioux Falls who was
present throughout the planning and development of the Center told us
that after the plan was approved and officials looked more closely at
their needs, they decided that the building planned for renovation
would not be large enough.  Factors that apparently contributed to
the miscalculation were that (1) no facilities experts were involved
in the site selection until after the plan was approved and (2) after
the plan was approved, Pierre and Mobridge, SD, were expected to
eventually be included in the consolidation, thereby calling for a
larger facility than was provided for in the plan.  At the conclusion
of our work, postal officials said that they were reevaluating
whether other communities would be brought into the consolidation. 

Transportation costs also exceeded projected amounts.  The plan
estimated that annual truck transportation costs, excluding the
one-time indemnity cost of cancelling transportation contracts, would
decrease by about $73,000 because of the consolidation.  Instead,
they increased by about $538,000 annually.  One reason, according to
a Postal Service transportation official, was because Aberdeen was
dropped from the consolidation, the Service had to maintain
Aberdeen's overnight network as well as the Dakota Central network. 
Also, price increases that boosted highway transportation contract
amounts were not recognized in the plan.  The unrecognized increases
in truck transportation costs are somewhat offset by the fact that
projected air transportation costs of about $382,000 annually were
not being incurred.  The plan expected that Dakota Central would be
linked to Minneapolis by daily air service connecting with the
Service's Express Mail network and commercial flights.  This service
is not a part of the current Dakota Central operation. 

Workhours and the number of personnel assigned to mail processing and
customer service significantly increased in the area after Dakota
Central opened.  We compared the number of employees assigned to this
work in the Huron, Watertown, and Mitchell post offices before the
consolidation to the amounts experienced by Dakota Central during the
comparable period after it was fully operational.  Specifically, we
compared postal accounting periods 1 through 5 of fiscal year 1993
(mid-September 1992 through early February 1993) for the three post
offices with the same periods in postal fiscal year 1994 for Dakota
Central, which opened in July 1993.  We excluded amounts estimated by
postal officials for Dakota Central work that was transferred from
Sioux Falls. 

Employment and workhours both increased by about 44 percent.  While
overall employment and workhours were reduced at the post offices,
they increased much more than expected at Dakota Central.  For
example, the plan provided for 24 nonsupervisory employees at the
Center, but 56 were assigned in 1994. 

The increase does not appear to be entirely attributable to an
increase in mail volume.  Differences in the way mail is counted
before and after the consolidation preclude a precise comparison of
before and after volumes.  Postal officials said that there were no
major new mailers or mail recipients in the area.  To some extent,
the exclusion of Aberdeen from the consolidation may have prevented
the Center from receiving as much mail volume as planned, thereby
causing the Center to be less efficient than expected.  However, this
factor alone would not appear to account for such a significant
increase in the number of mail processing employees and workhours. 


   STATUS OF PRIOR RECOMMENDATIONS
   AND NEED FOR POSTIMPLEMENTATION
   REVIEW
------------------------------------------------------------ Letter :6

In our previous report on Dakota Central, we recommended that the
Postal Service take steps to avoid or minimize service problems
associated with the consolidation.  Subsequently, we made additional
recommendations to address service and cost problems with a similar
consolidation in Waterloo, IA.\5 Responding to recommendations in
these two reports, the Postmaster General said that new guidelines
were being developed for planning and evaluating area mail processing
consolidations.  He said that the new guidelines would require
postimplementation reviews for each consolidation.  In late 1992,
postal headquarters began developing the new guidelines.  At the
conclusion of our work in South Dakota, the guidelines were still
being tested, and no target date had been set for their issuance. 

We believe that a postimplementation review of the Center would be
beneficial because (1) service is a major ongoing concern among
customers in the South Dakota area, (2) costs appear to have exceeded
the amounts expected, and (3) the Service is considering bringing
other communities under Dakota Central.  However, because the area
mail processing concept is in suspension and there are no guidelines
requiring postimplementation reviews, no requirement exists for a
review of the Center's effectiveness. 


--------------------
\5 See Postal Service:  Planned Benefits of Iowa Automated Mail
Facility Not Realized (GAO/GGD-94-78, Apr.  8, 1994). 


   RECOMMENDATION TO THE
   POSTMASTER GENERAL
------------------------------------------------------------ Letter :7

We recommend that the Postmaster General require that a
postimplementation review of Dakota Central be made to ensure that it
is achieving planned delivery performance and to more fully
understand why it is not achieving planned savings. 


   AGENCY COMMENTS
------------------------------------------------------------ Letter :8

The Postmaster General provided written comments on a draft of this
report, which are included in appendix II.  He said that the Service
initiated a postimplementation review of the Dakota Central Center
for performance and savings and that the review should be completed
in early January 1995.  He also said that the Service planned to
issue revised area mail processing guidelines by the end of the year,
and they will contain a requirement for timely postimplementation
reviews of each mail processing consolidation. 


---------------------------------------------------------- Letter :8.1

We are sending copies of this report to the South Dakota
congressional delegation and the Postmaster General.  Copies will
also be made available to others upon request. 

Major contributors to this report are listed in appendix III.  Please
call me on (202) 512-8387 if you have any questions about this
report. 

Sincerely yours,

J.  William Gadsby
Director, Government Business
 Operations Issues


DESCRIPTION OF MAIL TEST DESIGN
=========================================================== Appendix I

The mail test was designed to obtain data on the mail delivery times
within and among the various communities that were affected by the
Dakota Central consolidation.  Our test involved mailing 1,321 First
Class letters at various judgmentally selected locations within the
area and then determining the overnight delivery rates.  Our primary
interest was to determine whether overnight delivery had been reduced
within the principal communities as a result of the consolidation. 
Therefore, we sent slightly more than 75 percent of the letters
within the communities of Aberdeen, Huron, Mitchell, and Watertown,
SD.  We sent another 9 percent between these principal communities,
11 percent from outlying towns to the four principal communities, and
the remaining 5 percent between 8 remote towns.  Except for the four
principal communities, the selection of locations and numbers of
letters mailed was not intended to provide generalizable information. 

Within the principal communities, the number of letters mailed was
sufficient to provide 95-percent confidence that if the test were
repeated the results would be similar.  We sent all letters
separately so that they would be treated as single units rather than
as batched mail.  Batched mail would either have all been delivered
overnight or none would have been delivered overnight. 

  Mail within the principal communities of Huron, Aberdeen, Mitchell,
     and Watertown, SD:  We mailed letters (405 in Watertown, 270
     each in Aberdeen and Huron, and 61 in Mitchell) from various
     mail boxes and post offices in each community to a central
     address in the same town.  Because of an addressing error, we
     tested Watertown twice (270 in February and 135 in March).  We
     have no evidence to indicate that there were any significant
     differences between the two Watertown tests except for the
     addressing error. 

  Mail between the four principal communities of Aberdeen, Mitchell,
     Huron, and Watertown:  We mailed 123 letters from a single
     location in each community to a single address in each of the
     other 2 communities. 

  Mail from outlying towns to the four principal communities:  We
     mailed 144 letters from 12 smaller towns adjacent to the
     communities to a single address in the 4 communities. 

  Mail between remote towns:  We mailed 48 letters between 9 towns
     located near the end of mail routes originating from each of the
     4 principal communities. 

See figure I.1 for a map showing the location of these communities. 

   Figure I.1:  Location of
   Communities in the Mail Test

   (See figure in printed
   edition.)

   Source:  Postal Service.

   (See figure in printed
   edition.)

If we were to repeat our test within the four principal communities,
we can be 95-percent confident that the proportion of overnight
deliveries within Aberdeen would range between 100.0 and 98.9.  In
Huron, the similar range would be between 96.3 and 90.3.  In
Watertown, the range would be from 98.3 to 94.7.\\1

The selection of outlying and remote towns was judgmentally made to
cover the major delivery routes in the area and to give insight into
whether this mail experienced on-time or delayed delivery.  In
addition, the number of mailings was too small to permit the results
to be generalizable. 

We conducted our mail test in February and March 1994.  The results
of our mail test cannot be generalized to other weeks or months of
the year.  Mail processing and delivery performance can vary over
time due to changing mail volumes, weather conditions, employee
absence, equipment performance, and many other factors. 

Letters were deposited before the final scheduled pick up for the
day.  The mail was sent early in the workweek, and the results cannot
be generalized to the weekends.  In fact, the distribution process
changes on weekends when some mail is trucked to Sioux Falls, SD, for
processing. 



(See figure in printed edition.)Appendix II

--------------------
\1 Because of the problem encountered in regard to the Mitchell test,
we did not compute similar statistics for this community. 


COMMENTS FROM THE POSTAL SERVICE
=========================================================== Appendix I


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix III


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

Michael E.  Motley, Associate Director James T.  Campbell, Assistant
Director Leonard Hoglan, Assignment Manager


   DENVER FIELD OFFICE
------------------------------------------------------- Appendix III:2

Anthony R.  Padilla, Evaluator-in-Charge James R.  Moore, Evaluator
Cynthia Schilling, Reports Analyst Teri D.  Parham, Support
Technician
