Federal Prison Industries: Delivery Performance Is Improving but Problems
Remain (Letter Report, 06/30/98, GAO/GGD-98-118).
Pursuant to a congressional request, GAO: (1) developed and assessed
statistics on Federal Prison Industries' (FPI) delivery performance; and
(2) obtained the views of selected customer agencies' procurement
officials on FPI delivery practices.
GAO noted that: (1) FPI delivery performance is improving; (2) 8 of the
12 months in fiscal year (FY) 1997 had better on-time delivery
performance than the same months in FY 1996; (3) however, FPI fell short
of meeting its goal of 90 percent in FY 1997; (4) there was a wide
variation in FPI performance by customer agency and some variation by
product category; (5) the results of GAO's analysis and FPI's own
timeliness evaluations should be viewed with two caveats in mind; (6)
they both likely overstate timeliness because they did not account for
shipping time for orders with due dates specified as the day the order
should arrive at its destination; (7) GAO and FPI's timeliness
evaluations considered all shipments to be on-time if FPI data showed
that they left the factory on or before the due date; (8) accounting for
shipping time for orders with destination due dates would have improved
the accuracy of the timeliness evaluations and provided a better picture
of performance; (9) its review of 109 randomly selected orders showed
that over one-half of them had due dates in FPI's system that were later
than what customers had originally requested; (10) because of limited
documentation, GAO could not always determine the reasons due dates were
different, including whether FPI had not accepted them, or whether
customers were notified of the reasons for changes and approved of the
revised due dates; (11) although the results of these 109 orders were
not projectable to the universe of FPI orders, they raised questions
about which due dates should be used to measure timeliness, especially
from the customer's prospective; (12) just as GAO's analysis by customer
agency showed wide variation in FPI delivery performance, customer
agency officials within the Defense Logistics Agency, the Federal Supply
Service, Social Security Administration, and Department of Veterans
Affairs had mixed views on FPI's delivery performance, despite FPI's
goal to promote total customer satisfaction; (13) although GAO sought
the views of only selected customers, several key procurement officials
within these agencies were clearly dissatisfied with FPI's delivery
performance and practices; (14) FPI does not currently develop delivery
performance data by customer agency; and (15) without these data, FPI
was not in a good position to easily detect individual agencies'
problems with its performance and to improve overall customer relations.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: GGD-98-118
TITLE: Federal Prison Industries: Delivery Performance Is
Improving but Problems Remain
DATE: 06/30/98
SUBJECT: Correctional facilities
Customer service
Contractor performance
Management information systems
Contract monitoring
Delivery terms
Federal supply systems
Total quality management
******************************************************************
** 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
June 1998
FEDERAL PRISON INDUSTRIES -
DELIVERY PERFORMANCE IS IMPROVING
BUT PROBLEMS REMAIN
GAO/GGD-98-118
FPI Delivery Performance
(240267)
Abbreviations
=============================================================== ABBREV
BOP - Bureau of Prisons
CAL - Contractor Alert List
COED - customer order entry database
DLA - Defense Logistics Agency
DOD - Department of Defense
DOJ - Department of Justice
DSCP - Defense Supply Center Philadelphia
DSCC - Defense Supply Center Columbus
DSCR - Defense Supply Center Richmond
FAR - Federal Acquisition Regulation
FPI - Federal Prison Industries
FSS - Federal Supply Service
GSA - General Services Administration
GSAR - General Services Administration Acquisition Regulation
HHS - Department of Health and Human Services
NFC - National Furniture Center
OMB - Office of Management and Budget
SSA - Social Security Administration
SSA/OCG - SSA/Office of Operations, Contracts, and Grants
SSA/PM - SSA/Office of Property Management
VA - Department of Veterans Affairs
VHA - Veterans Health Administration
Letter
=============================================================== LETTER
B-278134
June 30, 1998
The Honorable Peter Hoekstra
The Honorable Mac Collins
The Honorable Roscoe G. Bartlett
The Honorable Nathan Deal
House of Representatives
This report responds to your July 8, 1997, request for information on
whether Federal Prison Industries (FPI) delivers its products and
services in a timely manner. Our objectives were to (1) develop and
assess statistics on FPI delivery performance and (2) obtain the
views of selected customer agencies' procurement officials on FPI
delivery practices. As agreed with your offices, our work for the
first objective involved developing information on how often FPI met
delivery due dates during fiscal years 1996 and 1997, including an
analysis by customer agency and product category. As part of this
work, we considered FPI's efforts to track delivery performance and
its practices with regard to setting due dates. Our work for the
second objective primarily involved holding discussions with selected
customer agency officials--from the Defense Logistics Agency (DLA),
General Services Administration's Federal Supply Service (FSS),
Social Security Administration (SSA), and Department of Veterans
Affairs (VA)--about their experiences with FPI delivery performance,
the due-date setting process, their efforts to monitor FPI delivery
performance, and available remedies for late deliveries. You
requested this work because of concerns you had about FPI's ability
to meet customers' due dates and its lack of incentive to do so.
Because of its status as a mandatory source supplier, agencies are
required to buy FPI products.
BACKGROUND
------------------------------------------------------------ Letter :1
FPI is a wholly owned government corporation managed by the
Department of Justice's (DOJ) Bureau of Prisons (BOP). FPI was
created by Congress in 1934 and serves as a means for managing,
training, and rehabilitating inmates. Under the trade name UNICOR,
FPI markets about 150 types of products and services to federal
agencies. Products include furniture, textiles, and electronic
components. Services include data entry, engine repair, and
furniture refinishing. In fiscal years 1996 and 1997, FPI had net
sales of about $496 and $513 million, respectively, in products and
services. In addition to buying products directly from FPI, agencies
buy FPI products from central supply agencies like FSS or, in the
case of the military, from DLA. FSS and DLA stock some FPI products
for sale to customers; FPI delivers other products directly to
customers when orders are placed with FSS or DLA. According to FPI,
agencies generally deal directly with FPI when procuring services.
When making orders for FPI products and services, agencies typically
send FPI a hard copy order; transmit the order electronically,
including through the Internet; or place the order by telephone.
To attain certain public policy objectives, federal law provides for
certain exceptions to the full and open competition requirement that
governs most acquisitions. One of these exceptions is set forth in
18 U.S.C. 4124, which provides that federal agencies are generally
required to purchase FPI products if they meet the buying agency's
requirements, and the prices charged do not exceed current market
prices. Subpart 8.6 of the Federal Acquisition Regulation (FAR)
implements this statutory provision. If FPI cannot meet the buying
agency's requirements, the FAR allows agencies to seek waivers from
FPI.\1 FPI's mandatory source status for products does not apply to
the services that it offers to its federal customers.
Over the years, supporters and critics of FPI have debated FPI's
mandatory source status and whether FPI provides quality goods, at a
reasonable price, and on a timely basis. Both sides recognize that
FPI has a social objective to manage, train, and rehabilitate inmates
through work programs. However, some critics have questioned whether
FPI's products and services have satisfied federal customers in terms
of timeliness, as well as quality and price, and whether FPI's
mandatory status results in unfair competition with the private
sector. In recent years, FPI has placed increased emphasis on
timeliness as a performance indicator, as well as on overall customer
satisfaction.\2 In its fiscal year 1996 operating plan, one of FPI's
eight long-term strategic goals was to "promote total customer
satisfaction by being competitive in marketplace price, quality,
customer service, and delivery standards." One of its corporate
objectives for fiscal year 1996 was to "meet or exceed all customer
requirements for price, quality, delivery, and service." Beginning
with its fiscal year 1997 operating plan, FPI established a specific
goal for on-time delivery performance, stating that it would publicly
recognize factories that reached 90 percent. For fiscal year 1998,
FPI increased its on-time performance goal to 98 percent. FPI
officials viewed these goals as ambitious; however, they said that
setting them represented an important step toward focusing on the
need to improve delivery performance.
--------------------
\1 Because FPI is a mandatory source provider for items it produces,
FPI's customers are required to obtain FPI's written authorization
prior to placing an order for a similar item through outside sources.
Federal agency customers can request a waiver by mail,
electronically, or by facsimile. The waiver request asks the
customer to justify that the product does not meet the basic needs of
the agency. Requests are to be evaluated on a case-by-case basis by
FPI sales consultants, but systems furniture waivers are handled by
FPI's systems projects group. Waivers that are denied can be
appealed to the FPI ombudsman in Washington, D.C.
\2 In March 1998, we reported that FPI did not have a systematic or
structured process for collecting and analyzing customer satisfaction
data so that conclusions can be drawn about customer satisfaction.
See Federal Prison Industries: Limited Data Available on Customer
Satisfaction (GAO/GGD-98-50, Mar. 16, 1998).
RESULTS IN BRIEF
------------------------------------------------------------ Letter :2
FPI delivery performance is improving--our work showed an upward
trend in the latter part of fiscal year 1997. In fact, our analysis
showed that 8 of the 12 months in fiscal year 1997 had better on-time
delivery performance than the same months in fiscal year 1996. In
addition, FPI had data to track timeliness and an approach for
evaluating results that showed improvements in fiscal year 1998.
However, our work also showed that FPI fell short of meeting its goal
of 90 percent in fiscal year 1997. Also, there was a wide variation
in FPI performance by customer agency and some variation by product
category. Specifically, our analysis showed that FPI shipped 72 and
76 percent of customer orders on or before the due date in fiscal
years 1996 and 1997, respectively. In 1997, on-time performance
ranged from 47 percent for VA to 92 percent for the federal
judiciary. In the same year, factories producing systems furniture
were the most timely at 85 percent, and factories producing
electronics were the least timely at 73 percent.
The results of our analysis and FPI's own timeliness evaluations
should be viewed with two caveats in mind. First, they both likely
overstate timeliness because they did not account for shipping time
for orders with due dates specified as the day the order should
arrive at its destination. Due dates specified by customers can be
origin--meaning that the orders should be shipped by the due date--or
destination, meaning that the orders should arrive at the customer's
location by the due date. FPI did not have data or a workable
approach that would allow us to account for shipping time for orders
with destination due dates. As a result, our analysis, as well as
FPI's own evaluations of timeliness, considered all shipments to be
on time if FPI data showed that they left the factory on or before
the due date. Accounting for shipping time for orders with
destination due dates would have improved the accuracy of the
timeliness evaluations and provided a better picture of performance.
Second, our review of 109 randomly selected orders showed that over
one-half of them had due dates in FPI's system that were later than
what customers had originally requested. Arriving at mutually
agreeable due dates may involve negotiation and compromise between
the supplier and the customer, whether the supplier is FPI or a
commercial vendor, because customer-requested due dates cannot always
be met. In addition, FPI's Chief Operating Officer said that one
reason due dates in FPI's system could have been later than what
customers originally requested was that FPI may have determined that
the customers' due dates were inconsistent with FPI production
capabilities at the time the order was placed, not accepted the
customer-requested due dates, and revised them to reflect production
lead times. However, because of limited documentation, we could not
always determine the reasons due dates were different, including
whether FPI had not accepted them, or whether customers were notified
of the reasons for changes and approved of the revised due dates.
Although the results for these 109 orders were not projectable to the
universe of FPI orders, they raised questions about which due dates
should be used to measure timeliness, especially from the customer's
perspective. FPI officials said that they could do a better job of
documenting the reasons due dates that customers request cannot be
met and whether the customers were notified of the reasons for
changes and approved of the new due dates.
Just as our analysis by customer agency showed wide variation in FPI
delivery performance, customer agency officials within DLA, FSS, SSA,
and VA had mixed views on FPI's delivery performance, despite FPI's
goal to promote total customer satisfaction. For example, these
views ranged from officials at FSS' National Furniture Center and VA
saying that FPI's performance was very poor and that FPI was very
difficult to deal with; to DLA officials, who said they were
generally pleased with FPI delivery performance. Although we sought
the views of only selected customers, several of the key procurement
officials within these agencies were clearly dissatisfied with FPI's
delivery performance and practices. In addition, we noted during our
review that FPI does not currently develop delivery performance data
by customer agency. Without these data, FPI was not in a good
position to easily detect individual agencies' problems with its
performance and to improve overall customer relations.
SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :3
To meet our first objective, we obtained and analyzed computerized
files from FPI's customer order entry database (COED) for fiscal
years 1996 and 1997, involving over 140,000 agency orders. FPI uses
COED, which is maintained at its Customer Service Center in
Lexington, KY, to track and record information on agency orders for
products and services. Agency orders can contain one to several
hundred line items. Each line item identifies a specific product and
quantity and has a distinct due date. Large orders with many line
items can have multiple products and due dates. Because agency
orders can vary tremendously, we generally followed the same approach
that FPI uses to define orders when it evaluates timeliness.
Specifically, we defined an order as a shipment of products or
provision of services expected by a customer for a given order on a
given day. We used the COED data to determine whether orders,
defined as such, were shown as being shipped in full by their
respective due dates.
We also examined how FPI accounts for shipping time for those orders
for which the due date is when deliveries should arrive at the
customer's location, not the date by which they should be shipped.
To perform a limited reliability assessment of the COED data, we
obtained hard copies of, and other available documentation for, 109
randomly selected agency orders and compared various data items in
the orders with data in COED. We also obtained and analyzed the
results of FPI's evaluations of timeliness for fiscal year 1997 and
the first half of 1998. We compared FPI's 1997 results with ours and
reconciled any differences. We did not independently verify FPI's
analysis for fiscal year 1998. In doing our work, we held many
interviews and discussions about FPI delivery practices with FPI
staff at the Customer Service Center in Lexington, KY, and at FPI
headquarters in Washington, D.C. We did not determine the reasons
individual agency orders were delivered late or assess the effect of
individual late deliveries on federal agencies.
To meet the second objective, we interviewed key procurement
officials from four agencies that are among FPI's largest buyers:
the General Services Administration's (GSA) Federal Supply Service
(FSS) in Arlington, VA; the Defense Logistics Agency (DLA) in Ft.
Belvoir, VA; the Social Security Administration (SSA) in Baltimore,
MD; and the Department of Veterans Affairs (VA) in Washington, D.C.
Collectively, these agencies accounted for over one-quarter of FPI's
1996 sales of $496 million. DLA was the second largest buying
component within the Department of Defense (DOD) next to the Army.
The other three agencies were among the top four civilian agencies
that buy from FPI.
In addition to contacting headquarters procurement officials at these
agencies, we also contacted one of GSA's nationwide commodity
centers, FSS' National Furniture Center (NFC), and three of DLA's
nationwide supply centers: the Defense Supply Center Philadelphia
(DSCP) in Philadelphia, PA; the Defense Supply Center Columbus (DSCC)
in Columbus, OH; and the Defense Supply Center Richmond (DSCR) in
Richmond, VA. Within SSA, we spoke with officials from the two
buying components that purchase products from FPI: the Office of
Property Management (SSA/PM) and the Office of Operations, Contracts,
and Grants (SSA/OCG). Within VA, we spoke with officials from the
Office of Administration and the Office of Acquisition and Materiel
Management, as well as officials with the Veterans Health
Administration (VHA) from the offices of Patient Care Services and
Environmental Management Services.
We obtained information and views from these officials on several
topics related to timeliness, including how FPI performed in meeting
due dates, how due dates were set, whether due dates were driven by
FPI production capabilities or mission needs, what processes were
used to monitor FPI delivery performance and that of private vendors,
and what remedies were available in the event of a late delivery from
FPI and private vendors. Using FPI's waiver database, we also
determined the extent to which FPI granted customers waivers from
buying FPI products for reasons related to timeliness.
We did our work between July 1997 and May 1998 in accordance with
generally accepted government auditing standards. Appendix II
contains a more detailed discussion of our objectives, scope, and
methodology. We requested comments on a draft of this report from
the Director, Bureau of Prisons (BOP). These written comments are
discussed near the end of this letter and are reprinted in appendix
III. We also held exit conferences with program officials of the
customer agencies we visited to verify applicable data, facts, and
opinions presented in this report.
TIMELINESS STATISTICS AND THEIR
LIMITATIONS
------------------------------------------------------------ Letter :4
Our analysis using FPI data and its general approach for measuring
timeliness showed that FPI shipped 72 and 76 percent of customer
orders on or before the due date in fiscal years 1996 and 1997,
respectively. It also showed a wide variation in FPI performance by
customer agency and some variation by product category. FPI did not
meet its 90 percent on-time delivery goal in any of the months in
1997 or in any of the product categories for the year as a whole.
However, FPI's performance improved toward the end of 1997, reaching
87 percent for the month of July, and its own data showed that
timeliness reached 89 percent in March 1998. In addition, our work
showed that the average amount of time it took FPI to ship products
in all but one of the product categories decreased from 1996 to 1997.
These statistics provide insights into FPI delivery performance, such
as the variation by product category and customer agency. However,
it is important to recognize that they likely overstate timeliness
because they did not account for shipping time for orders with due
dates specified as the day the order should arrive at its destination
and not the day by which the order should be shipped. As a result,
our analysis, as well as FPI's evaluations of timeliness, considered
all shipments to be on time if they left the factory on or before the
due date. In addition, our review of 109 randomly selected orders
showed that over one-half of them had due dates in FPI's system that
were later than what customers had originally requested. Because of
limited documentation, we could not always determine the reasons due
dates had changed or whether customers were notified of the reasons
for the changes and approved of them. Furthermore, although the
results for these 109 orders were not projectable to the universe of
FPI orders, they raised questions about which due dates should be
used to measure timeliness, especially from the customer's
perspective. If customers are not satisfied with due dates or if
this limitation is not reflected in the analysis, the usefulness of
any timeliness evaluations that rely on them comes into question.
STATISTICS ON TIMELINESS
---------------------------------------------------------- Letter :4.1
FPI shipped 72 and 76 percent of customer orders on or before their
due dates in fiscal years 1996 and 1997, respectively. FPI improved
its performance in the latter months of fiscal year 1997, reaching 87
percent for July 1997. In fact, our analysis showed that 8 of the 12
months in fiscal year 1997 had better on-time delivery performance
than the same months in fiscal year 1996. It should be noted,
however, that timeliness did decrease to 84 and 82 percent,
respectively, for August and September of 1997, and FPI did not meet
its 90 percent on-time delivery goal in any of the months in 1997.
For the 2-year period, timeliness ranged from a low of 59 percent in
November 1995 to a high of 87 percent in July 1997. On average,
6,663 orders were due in each month of 1996 and 1997, ranging from
5,057 in February 1996 to 8,227 in September 1997. These orders
covered the range of products FPI manufactures, as well as services,
such as engine repair, data entry, and furniture refinishing.
Figures 1 and 2 and tables I.1 and I.2 in appendix I show the results
of our analysis by month for fiscal years 1996 and 1997.
Figure 1: FPI Timeliness for
Fiscal Year 1996
(See figure in printed
edition.)
Source: GAO analysis of FPI data.
Figure 2: FPI Timeliness for
Fiscal Year 1997
(See figure in printed
edition.)
Source: GAO analysis of FPI data.
FPI's six major product categories--data graphics,\3 electronics,
furniture, metals, systems furniture, and textiles--showed some
variation in delivery performance. Our analysis showed that
factories producing metals--which include metal shelving, lockers,
and storage cabinets--were the most timely in 1996 at 86 percent.
Factories producing furniture--which includes desks, bookcases, and
ergonomic chairs--were the least timely in 1996 at 66 percent. In
1997, factories producing systems furniture were the most timely at
85 percent, and factories producing electronics--which include
cables, connecters, and circuit assemblies--were the least timely at
73 percent. Median production times in 1996 ranged from 23 days for
data graphics factories to 102 days for factories producing
furniture. In 1997, production times ranged from 14 days for data
graphics factories to 89 days for factories producing furniture. For
1997 as a whole, none of the averages for the six product categories
met FPI's 90-percent goal. However, timeliness in four of the six
product categories improved from 1996 to 1997; and production time
decreased in every product category except metals, where it increased
slightly. Table I.3 in appendix I shows the results of our analysis
by product category for 1996 and 1997, as well as some examples of
the types of products under each category.
Our analysis by customer agency showed wide variations in
performance. To examine the data by agency, we focused on the top
buyers of FPI products according to FPI's fiscal year 1996 sales
report. Orders shipped on or before the due dates to different
customer agencies ranged from a low of 47 percent for VA orders to a
high of 88 percent for federal judiciary orders in 1996. In 1997, VA
and the federal judiciary were again the lowest and highest at 47 and
92 percent, respectively. Table I.4 in appendix I shows the results
of our analysis by customer agency for 1996 and 1997.
In addition, we analyzed products from FPI's "quickship" program.
FPI guarantees that it will ship products in this program within 30
days of receipt of agencies' orders. Products in FPI's quickship
catalogue include certain types of ergonomic chairs and other
furniture, linens, clothing, targets, and traffic signs. Despite
FPI's guarantee that it will ship these products within 30 days, we
found that only 75 and 70 percent of quickship items were shipped
within 30 days or less in fiscal years 1996 and 1997, respectively.
Our analysis also showed that only 79 and 63 percent of the due dates
for quickship products that were in FPI's system for fiscal years
1996 and 1997, respectively, were within the 30-day time frame.
As mentioned before and explained in detail in appendix II, we
generally followed the approach FPI uses to define orders when
evaluating timeliness, because agency orders vary tremendously, often
involving multiple products and due dates. On-time delivery was one
of several performance indicators--which include sales and earnings,
inventories, and inmate staffing--that FPI updated on a regular
basis. FPI monitored on-time delivery performance on a monthly basis
and by product category and factory. However, we noted during our
review that in evaluating timeliness, FPI did not track performance
by customer agency. The monthly trend shown in our analysis for 1997
(fig. 2) generally matched FPI's results. However, FPI's results
were slightly higher--by an average of just over 2 percent for the
months in 1997. On the basis of discussions with FPI staff, we
determined that this was likely due to two minor differences between
our analysis and FPI's.
First, as is discussed in appendix II, we did not group the orders,
as we defined them, by factory as FPI does for its evaluations,
because we were not assessing the performance of individual factories
from month to month. The effect of this added grouping was that the
base number of orders FPI used was higher, and slightly more of these
additional orders were counted as being on time rather than late.
This accounted for most of the difference between the two analyses.
Second, FPI's analysis included blanket orders, which have regular
shipments over an extended period of time. These orders make up
about 3 percent of the orders, but they involve a large volume of
data. Early in our review, we decided to request data only on
regular orders, excluding the blanket orders, to minimize the amount
of data we were requesting, which ultimately included over 600,000
line items.
For the first half of fiscal year 1998, FPI's timeliness statistics
showed that about 85 percent of the orders were on time in October
and November. In December and January, the percentages went down to
79 and then 78 percent and then increased to 83 and 89 percent in
February and March, respectively. According to FPI officials,
timeliness reached 90 percent in April 1998. As mentioned before,
FPI set a goal of reaching 98 percent on-time delivery in fiscal year
1998.
--------------------
\3 The data graphics category includes signage, printing, and data
entry, as well as some other miscellaneous products and services,
such as optics, brushes, and laundry services.
STATISTICS DID NOT ACCOUNT
FOR SHIPPING TIME
---------------------------------------------------------- Letter :4.2
A limitation of our analysis and FPI's timeliness evaluations is that
they did not account for shipping time for due dates specified by
customers as the day the order should arrive at its destination. Due
dates specified by customers can be origin--meaning that the orders
should be shipped by the due date--or destination, meaning that the
orders should arrive at the customer's location by the due date. FPI
had data on estimated shipping times for only 18 of its 96 factories
that we could have used in our analysis, and we found that several of
the other factories received orders with destination due dates.
Estimated shipping times for the 18 factories ranged from 3 to 10
days. In addition, FPI staff who wrote FPI's computer program for
evaluating timeliness said they had problems developing a workable
approach for accounting for shipping time where appropriate. In the
absence of data or a workable approach for incorporating shipping
time, our analysis, as well as FPI's timeliness evaluation,
considered all shipments to be on time if FPI's data showed that they
left the factory on or before the due date.
The FPI official who compiled the performance indicators acknowledged
that the inability to account for shipping lead times was a fault in
FPI's system. This official said that FPI's efforts to measure
timeliness have always been hampered by the fact that FPI never knows
exactly when the customer receives the products. The official said
that subsequent FPI efforts to estimate shipping time have been
unsuccessful. The official indicated that ideally, orders with
destination due dates could be flagged as they come in, and a
standard process could be used to estimate shipping time for
incorporation into the due date, depending on the type of order.
Because we could not account for shipping time for orders with
destination due dates, our analysis and FPI's timeliness evaluations
likely overstate timeliness because they treated every due date as an
origin date. Although we could not determine how much our statistics
on timeliness would decrease if we could account for shipping time
where appropriate, our work indicated that accounting for shipping
time would have some impact. First, although our data were not
projectable, we found that 53 of the 109 orders we randomly selected
to assess the data in FPI's system had due dates that customers had
specified as destination. Second, as mentioned before, available
data for 18 of the factories showed that estimated shipping times
ranged from 3 to 10 days; and about 16 percent of the 428,000 line
items that were on time in orders due in 1996 and 1997 were shipped
on 1 of the 2 days prior to the due date or on the due date itself.
In these cases, if the line items had destination due dates, they
likely would have been late given the minimum estimate of 3 days for
shipping time.
ORIGINAL CUSTOMER- REQUESTED
DUE DATES WERE BEING CHANGED
---------------------------------------------------------- Letter :4.3
Our review of 109 randomly selected orders showed that 84 of them had
due dates in FPI's order entry system that were different from what
customers had originally requested. Specifically, 63 of the orders
had due dates that were later than what customers originally
requested, and 16 had due dates that were earlier. Five orders
involved a combination of due dates for different line items that
were both later and earlier than what customers originally requested.
Differences between originally requested due dates and the later due
dates in FPI's system ranged from 2 to 333 days, and differences for
due dates that were earlier ranged from 1 to 58 days. Only 25 of the
109 orders had due dates in FPI's system that reflected what the
customers had originally requested.
Of the 68 orders with due dates that were later than what customers
requested--the 63 with dates that were later plus the 5 with a
combination of later and earlier due dates--available documentation
and information in FPI's system showed that 7 due date changes were
initiated or caused by the customer. Nineteen had due dates that
were later than what the customer had originally requested, because
FPI's production capacity was overloaded at the time the order was
placed, or FPI determined that the lead time the customer specified
was insufficient. In the remaining 42 cases, the documentation FPI
provided for the orders and information in its system did not explain
the reasons the due dates were later than what the customers had
originally requested. FPI researched the 42 orders and provided an
explanation for most of the differences. These explanations included
customers specifying insufficient lead times for production, FPI's
production capacity being overloaded at the time the order was
placed, and factory or project managers changing due dates for no
specified reasons. The officials indicated that in the orders for
which due dates were made earlier than the customer requested, FPI
most likely was able to provide the product sooner than the customer
anticipated.
FPI's Chief Operating Officer said that one reason due dates in FPI's
system could have been later than what customers originally requested
was that FPI may have determined that the customers' due dates were
inconsistent with production lead times at the time the order was
placed, not accepted the customer-requested due dates, and revised
them to reflect production lead times. In these instances, the Chief
Operating Officer said that FPI considered the revised due dates to
be the original due dates, not the customer-requested due dates that
were not accepted. In contrast, he said that a true due date change,
from FPI's perspective, would involve FPI accepting a customer order
and its accompanying due dates and then revising them after
acceptance. However, he pointed out that in such instances, the
original due dates should remain unchanged in FPI's system unless the
customer initiated the due date revision.
Officials at the Customer Service Center in Lexington, KY, pointed
out that FPI's production lead times and capabilities can vary
throughout the year. For example, in a peak time such as the end of
the fiscal year, production lead times for some product groups can
increase from 90 to 150 days. These officials said they made every
effort to meet customers' needs through negotiating due dates,
offering alternative products and discounts, and offering waivers
from buying FPI products if necessary. Furthermore, they said that
in many cases, they maintained close working relationships with
customers; and for large orders, field representatives from the
individual factories were often in contact with customers regularly.
Our review also showed that there was limited documentation or
information in FPI's system showing that customers had been notified
that the due dates they requested could not be met, the reasons for
the changes, and that customers approved of the changes. FPI
officials told us that it is their policy to notify the customer and
request approval if customer-requested due dates cannot be met,
although they added that if they received no response from the
customer, they assumed that the new dates they developed were
acceptable. As mentioned before, 68 orders had due dates that were
later than what customers requested--available information showed
that changes for 7 of these were initiated by or caused by the
customer. Of the 61 remaining orders, 23 had available documentation
or information in FPI's system showing that customers were notified
that the due dates they requested could not be met and the reason for
the changes.
In another 24 cases, FPI provided us with order acknowledgement
letters they said were sent to customers. These acknowledgement
letters identified the new due dates FPI had developed on the basis
of its production lead times and capabilities at the time the order
was placed. However, the acknowledgement letters did not provide the
customers with the reasons the due dates they requested could not be
met or ask for the customers' approval of the changes. In addition,
these acknowledgement letters gave no indication that FPI did not
accept the customers' requested due dates or that the dates were
inconsistent with FPI's production lead times. For the other 14
orders, FPI had no documentation or information in its system showing
that customers were notified that the due dates they requested could
not be met. For the 61 orders, documentation or information in FPI's
system showing that the customers had approved of the date changes
was available in only 11 of these cases.
In doing our work, we also found that in 16 of these 61 orders, FPI
was able to ship the orders by the original customer-requested due
dates, despite the due date changes that occurred. However, FPI
missed original customer-requested due dates in the rest of the
orders, except for one where we could not make a determination
because we could not identify the exact due date. In addition, FPI
missed revised due dates that were in its system in 24 of these
orders. The results of our work related to the 109 orders were not
projectable to the universe of agency orders; however, they provided
us with insights into the COED data, as well as into FPI delivery
practices.
We recognize, as FPI officials pointed out, that due dates initially
requested by customers may be inconsistent with production
capabilities and that a number of valid reasons can exist for due
date changes. In addition, FPI may encounter increases in workload
and other scheduling difficulties, as any supplier would, at times
when customers are under pressure to meet their agencies' mission
needs. As a result, arriving at mutually agreeable due dates may
involve negotiation and compromise between the supplier and the
customer, whether the supplier is FPI or a commercial vendor. This
is particularly so in orders placed with commercial vendors involving
indefinite delivery type contracts. FAR Sections 16.501-2 through
16.504 describe these types of contracts as having delivery schedules
not known at the time of contract award and deliveries to be
scheduled by placing orders with the contractor. Nonetheless, the
lack of documentation showing the reasons due dates in FPI's system
were different from what customers requested and whether customers
were notified of these reasons and approved of the changes impeded
our ability to assess the appropriateness of many of the due dates
for the sample of orders. Although the results for these 109 orders
were not projectable to the universe of all FPI orders, they raised
questions about which due dates should be used to measure timeliness,
especially from the customer's perspective. FPI officials said that
they could do a better job documenting why customer-requested due
dates could not be met and whether the customers were notified of the
reasons for changes and approved of the new due dates.
SELECTED CUSTOMER AGENCY VIEWS
ON FPI DELIVERY ISSUES
------------------------------------------------------------ Letter :5
Just as our analysis by customer agency showed wide variations in FPI
delivery performance by customer agency, officials within DLA, FSS,
SSA, and VA who were involved in buying FPI products had mixed views
on FPI's performance in meeting delivery due dates, as well as on
related FPI delivery practices. Headquarters officials in FSS and
DLA, officials at two of the DLA supply centers, and an official in
the Office of Property Management at SSA for the most part spoke
positively about FPI's overall delivery performance. In contrast,
officials from VA, FSS' National Furniture Center, one of the DLA
supply centers, and SSA's Office of Operations, Contracts, and Grants
had negative experiences. Officials from DLA, FSS, and one of the
two offices we contacted in SSA said that delivery due dates were
driven primarily by FPI production capabilities and not agency
mission needs--some of the officials expressed concern with this
practice. VA officials said that it seemed that FPI attempted to
work with them to reach mutually agreeable due dates; however, these
officials said that FPI generally performed poorly in meeting those
dates. The views expressed by these customer agency officials may
not be representative of the views of all FPI customers. However,
despite FPI's goal to promote total customer satisfaction, our work
showed that certain key agency customers were clearly dissatisfied
with FPI's delivery performance and practices.
These agencies--DLA, FSS, SSA, and VA--all said they monitored
timeliness while administering contracts for all vendors, including
FPI.\4 Although some components within these agencies had data on FPI
and commercial vendor delivery performance, none of the agencies had
overall data comparing FPI performance to that of commercial vendors.
Nonetheless, contracting officers' leverage in resolving procurement
problems was different for FPI from the leverage contracting officers
had for private sector vendors, because the rules that typically
govern contracts with commercial vendors do not apply to FPI.
However, agencies can use performance information to seek waivers
from FPI so they can buy products commercially and can seek remedies,
including damages for late deliveries.
--------------------
\4 In this report, we use the term "contract" to include agency
agreements with FPI.
VIEWS ON FPI DELIVERY
PERFORMANCE WERE MIXED
---------------------------------------------------------- Letter :5.1
As mentioned before, our analysis showed wide variation in FPI
performance by customer agency. Our discussions with selected
customer agency officials we contacted from DLA, FSS, SSA, and VA
also reflected this mixed FPI performance by agency. Headquarters
FSS officials said that FPI's delivery performance had generally
improved over the last 10 years and that overall, FPI met FSS' needs.
FSS did not have overall data comparing FPI delivery performance to
that of commercial vendors.\5 FSS officials did say, however, that
the workload associated with addressing delinquent FPI deliveries was
significant. Officials at FSS' National Furniture Center (NFC) in
Arlington, VA, had more significant problems and discussed the
increased workload and their general dissatisfaction with FPI
delivery performance. This commodity center acts as an agent for
federal customers who buy furniture--FPI's largest product
line--through GSA's special order and multiple award schedules
programs.
FSS/NFC officials said that FPI delivery performance was a big
problem for the center. They said that the additional workload
caused by having to address FPI delivery problems exceeded staff
resources available and that this was compounded by FPI's
nonresponsiveness, such as not returning phone calls. They believed
that FPI was not sensitive to the due dates it agreed to with
customers nor to the needs of customers ordering furniture. They
added that customers often did not want to acquire furniture from FPI
because of previous experiences when FPI was unreliable. FSS/NFC
staff provided us with data showing that for fiscal year 1997, FPI
delivered 69 percent of the 12,688 line items of furniture the center
ordered on time.
DLA headquarters officials, as well as officials at two of the three
supply centers we contacted, were generally pleased with FPI delivery
performance and generally thought it compared favorably with that of
commercial vendors. Although these officials said that DLA did not
have overall data comparing FPI performance to that of commercial
vendors, they said that DLA routinely stocked extra products, and the
officials did not perceive that their mission had been hampered by
FPI delivery performance. Despite this, officials from the Defense
Supply Center Philadelphia (DSCP) said that FPI was unable to meet
the needs of the Army in Bosnia for body armor. According to FPI
officials, FPI experienced difficulty in obtaining raw materials from
its vendors on these contracts. DSCP's Deputy Director of Clothing
and Textiles said that this example illustrated the danger of
allowing a mandated supplier to be the sole provider of a product
manufactured to military specifications. This official said that
commercial suppliers cease manufacturing products when this happens,
because they know federal agencies must buy from the mandatory source
and there is an insufficient civilian market for many military
products. Because of this experience, DSCP now has a back-up
supplier for body armor.
Data from two of the DLA supply centers we contacted--DSCP and the
Defense Supply Center Columbus (DSCC)--showed mixed FPI performance.
DSCP reported that FPI's overall delinquency rates on all products
were 29 and 21 percent in fiscal years 1996 and 1997, respectively.
The commercial supplier rates for this center were 8 and 5 percent,
respectively. DSCC provided data from electronics acquisitions that
showed that FPI performed better overall in timeliness and quality
combined than commercial vendors--98 out of 100 points--compared to a
score of 79 out of 100 for all commercial vendors. Defense Supply
Center Richmond (DSCR) officials did not have any data on FPI
performance, but they said that their experience had been generally
positive.
FPI's two main buyers within SSA--the Office of Operations,
Contracts, and Grants (SSA/OCG) and the Office of Property Management
(SSA/PM)--had differing views of FPI performance. SSA/OCG handles
small orders for furniture that are sent to FPI on an as-needed
basis, and SSA/PM handles mostly systems furniture orders that are
scheduled up to a year in advance. The Director of the Center for
Ergonomic Property in SSA/PM said that FPI had not missed a delivery
due date since 1993 and had sometimes exceeded the specified time
frames. On the other hand, the Director of Operations Contracts in
SSA/OCG told us that commercial vendors were much more responsive
than FPI and delivered products in a shorter time period. This
official indicated that although the effect of missed delivery dates
was fairly negligible, SSA/OCG was generally dissatisfied with the
requirement to do business with FPI because of problems with
timeliness, as well as other factors.
Although VA has not developed overall data on FPI delivery
performance, headquarters officials from the Office of Acquisition
and Materiel Management, as well as program officials with the
Veterans Health Administration (VHA), said that FPI's delivery
performance had been a continuing, significant problem. Products
with delivery problems have been furniture, seating, and signage for
VA's major renovation of its headquarters building and other
facilities, textile products for the day-to-day operating needs of
VHA medical centers, and patient care requirements for prescription
eyeware. For example, for the renovation of VA's headquarters
building between 1992 and 1996, VA requested a waiver for the entire
requirement from FPI because of past performance problems. FPI
granted the waiver for the systems furniture portion of this project
but denied the waiver for the case goods (such as credenzas and
bookcases), seating, and signage. VA contracting officials
responsible for this project said that a commercial supplier met the
delivery requirements for the systems furniture; however, FPI
delivery performance on the remaining requirements was poor. In
addition, they said that many of the products were defective and
unacceptable, and customer support was almost nonexistent.
According to the VHA officials we contacted, another area with
significant delivery problems was prescription eyeware for veterans.
According to an official with VHA Patient Care Services, many
commercial suppliers were able to provide prescription eyeglasses in
5 to 7 days; and a Navy optical laboratory, which meets DOD's needs
for eyeglasses, was able to provide them to military personnel in 4
to 6 days. An official with the American Optometric Association said
that wholesale laboratories now have an average turnaround time of
about 2 days for complete jobs. Under FPI's mandatory source status,
VA believed it was required to purchase eyeglasses from FPI, until an
opinion was issued by VA's Office of General Counsel in July 1997
concluding that FPI was not a mandatory source of supply for VA's
procurement of eyeglasses. VA's experience with FPI showed that the
private sector provided much more timely delivery. VA officials told
us that in 1993, they requested a blanket waiver from FPI so they
could buy eyeglasses commercially. FPI did not approve the blanket
waiver but said that individual medical centers could request
waivers, and FPI would review them on a case-by-case basis.
Consequently, numerous individual hospitals tracked FPI performance
and requested waivers. For example, this official said that the VHA
medical center in Pittsburgh, PA, monitored FPI deliveries over a
2-month period in 1996 to support requesting a waiver from FPI and
found that although FPI had promised the eyeglasses within 7 days of
receiving the order, only 17 percent were delivered in that time
period. In fact, he said that 32 percent of the deliveries took more
than 20 days.
In light of the problems VA has had with FPI delivery performance, as
well as product quality and pricing, an official with VA's
Environmental Management Service said that VA had submitted a
legislative proposal to the Office of Management and Budget (OMB)
each year since 1988 to exempt VA from FPI's mandatory source
requirement for certain products. OMB has not approved any of these
proposals. In July 1997, VA's Office of General Counsel issued an
opinion that FPI is not a mandatory source for VA's purchase of
eyeglasses. The opinion concluded that VA could procure eyeglasses
from the commercial marketplace because 38 U.S.C. 8123 allows VA to
procure prosthesis devices in a manner the Secretary determines
proper without regard to any other provision of law and VA has
defined eyeglasses as a prosthesis device. According to FPI's Chief
Operating Officer, FPI did not dispute this decision; therefore, VA
may now purchase eyeglasses commercially. FPI officials said that
seven VA hospitals were still buying eyeglasses from FPI and provided
data indicating FPI's on-time delivery for these hospitals, using a
7-day delivery standard, ranged from 71 to 86 percent for the period
October 1, 1997, through May 21, 1998.
--------------------
\5 We reported in March 1998 (GAO/GGD-98-50, Mar. 16, 1998) that
both commercial vendors and FPI contracts were on FSS' Contractor
Alert List (CAL), which identifies contractors that fail to meet some
portion of their contract terms and are considered to be a higher
risk to the government as future contractors. FSS officials told us
that as of August 1997, 17 of FPI's 60 contracts with FSS were on the
CAL because of timeliness problems. Being placed on the CAL for
timeliness problems means that less than 90 percent of the orders for
a contract were on time.
DUE DATES OFTEN REFLECTED
FPI PRODUCTION CAPABILITIES
---------------------------------------------------------- Letter :5.2
As discussed earlier, FPI officials said that due dates were often
driven by FPI's production lead times. Officials from DLA, FSS, and
one of the two offices we contacted in SSA also said that delivery
due dates were driven primarily by FPI production capabilities and
not agency mission needs. Headquarters FSS officials pointed out
that its General Services Administration Acquisition Regulation
(GSAR) specifically states that "contracting officers shall establish
delivery schedules based on the lead time required by FPI."\6 These
FSS officials said that due dates were set on a case-by-case basis
and usually involved working with FPI to reach a mutual agreement.
They said that when contracting with FPI and developing due dates,
their approach was to weigh heavily on FPI's ability to deliver. One
top FSS official indicated that from a practical standpoint, it would
not make sense to set due dates that FPI factories could not meet.
Although the headquarters FSS officials indicated that FSS worked
with FPI within its production capabilities, FSS/NFC officials viewed
the due date setting process with FPI as more problematic. FSS/NFC's
officials said that for certain furniture products, FPI dictated the
due dates to FSS/NFC staff. They said that agencies often came to
FSS/NFC for contracting support when purchasing furniture from FPI.
FSS/NFC staff tried to negotiate due dates with FPI, but the
officials said that FPI's proposed due dates rather than the
customers' were generally used in the contracts. They said that
better prices, better quality furniture, and better delivery times
were often available through commercial suppliers on FSS' schedules.
DLA headquarters officials said that the due dates were driven mostly
by FPI production capabilities, but they did not view this as a
problem. For the most part, these officials said that DLA planned
its orders well in advance and often stocked extra levels. They said
that DLA inventory managers and DLA's automated systems alerted
contracting officers when it was time to make an order to FPI so that
future needs would be met. These officials said that there were
exceptions to this practice, such as when a critical need arose. In
these cases, they negotiated due dates with FPI. Although DLA
officials focused on the fact that they stock items and are able to
work within FPI's production capabilities, we have reported on
several best practices that have been successfully used in the
private sector to reduce inventory levels and logistics costs.\7 In
general, these practices provide inventory users with a capability to
order supplies as they are needed and then to deliver those items
directly to the customer within hours after the order is placed. DLA
has successfully used this approach in some areas. If DLA begins
relying more on these practices, it would seem that the due dates it
sets with its suppliers--including FPI--would take on greater
importance.
The Director of the Center for Ergonomic Property in SSA/PM was the
only official we contacted who said that due dates were driven by
agency mission needs. This official said that the delivery schedules
are set to accommodate SSA installation needs and that FPI always
meets those due dates. On the other hand, the Director of Operations
Contracts in SSA/OCG said that due dates were mostly dependent on FPI
production capabilities and less so on the particular need of the
requesting office within SSA. This official added that on the basis
of past experience, SSA/OCG expected that FPI would miss the due
dates that FPI itself established. For example, this official said
that on August 28, 1996, SSA/OCG placed an order for 83 chairs and
agreed with FPI to a due date of November 15, 1996. SSA/OCG then
received a notice from FPI on September 6, 1996, indicating that it
was unable to obtain the raw materials needed to produce the chairs
and that the new due date was December 3, 1996, to which they agreed.
FPI was then unable to ship the chairs until January 11, 1997. The
SSA/OCG official added that missed due dates often involved partial
shipments, in which most of the order arrived on time but some small
portion was late. This official also said that FPI often would
change due dates, and the procurement officer would not find out
about the changes until after the requesting office did not receive
the expected delivery.
VA officials did not have a view on whether due dates were primarily
driven by production capabilities or agency mission needs. VA
officials said that it seemed that FPI attempted to work with them to
reach mutually agreeable due dates. However, they said that FPI
generally performed poorly in meeting those due dates, which our
analysis of FPI's data confirmed with regard to VA.
As mentioned before, we recognize that arriving at mutually agreeable
due dates may involve negotiation between the supplier and the
customer, whether the supplier is FPI or a private vendor. In
addition, the views expressed by the agency officials we contacted
about the due date setting process, as well as FPI's delivery
performance overall, may not be representative of the views of all
FPI customers. However, meeting agencies' due dates is important for
customer satisfaction, and our work showed that certain key agency
customers were clearly dissatisfied with FPI delivery performance and
practices. Also, as mentioned before, FPI monitored on-time delivery
performance on a monthly basis and by product category and factory;
however, it did not monitor performance by customer agency. This may
have made it difficult for FPI to recognize when certain agencies or
agency components were experiencing a low rate of timeliness with FPI
deliveries.
--------------------
\6 This requirement, in GSAR at 48 C.F.R. Part 508, entitled
"Required Sources of Supplies and Services," applies only to GSA.
The other agencies, DLA, SSA, and VA, do not have a similar
requirement in their acquisition regulations.
\7 Inventory Management: Greater Use of Best Practices Could Reduce
DOD's Logistics Costs (GAO/T-NSIAD-97-214, July 24, 1997).
AGENCIES' USE OF PERFORMANCE
INFORMATION AND THE WAIVER
PROCESS
---------------------------------------------------------- Letter :5.3
DLA, FSS, SSA, and VA all said they monitored timeliness while
administering contracts for all vendors, including FPI. Although
some components within these agencies had data on FPI and commercial
vendor delivery performance, none of the agencies had overall data
comparing FPI performance to that of commercial vendors.
Nonetheless, as we reported in March 1998, contracting officers'
leverage in resolving procurement problems was different for FPI from
the leverage the contracting officer had for private sector vendors,
because the rules that typically govern contracts with commercial
vendors do not apply to FPI.\8 Further, when agencies develop varying
types of information on current and past performance of vendors,
including FPI, there are major distinctions in how agencies can use
this information.
Specifically, for commercial vendors, agencies can use past
performance information showing timeliness problems as a factor to
consider in the award of contracts. However, for FPI, agencies
cannot use this information to deny awarding a contract to FPI
because, under the law, FPI is a mandatory source of supply.
Agencies can, however, use this information to negotiate with FPI on
delivery time frames or to seek a waiver from FPI so they can buy
from a commercial vendor that can better meet their delivery
requirements. As with awarding contracts, agencies cannot use
current performance information to cancel or terminate an existing
contract with FPI unless agencies request cancellation or termination
provisions during the negotiation process and FPI agrees to include
them in the contract. According to FPI, agencies can, however, use
current performance information to seek other remedies against FPI,
including damages for late deliveries. We did not determine the
extent to which agencies seek, or FPI agrees to authorize, these
remedies; nor did we compare or evaluate the use of remedies by
contracting officers in connection with contracts with FPI or
commercial vendors.
Regarding waivers for reasons related to timeliness, our analysis of
FPI waiver data showed that 36 and 29 percent of the 21,900 and
24,300 line items for which FPI had granted waivers in 1996 and 1997,
respectively, were because FPI could not meet customers' time frames.
Inability to meet customers' time frames was the most common reason
FPI granted waivers in these years. FPI also provided data showing
that it approved 85 and 83 percent of the line items for which
customers requested waivers for any reason in fiscal years 1996 and
1997, respectively. Our discussions with agency officials showed
that they generally do not seek waivers for reasons related to
timeliness, with the exception of VA. FSS officials said that their
staff typically did not request waivers but instead worked with FPI
to find an acceptable solution short of requesting a waiver. This
may have involved extending the due dates for orders or requesting
that FPI change the production point. An FSS official said that
using this approach, although resource intensive, was often in the
best interest of the customer, because the customer would get the
product sooner than if FSS had to obtain the waiver and initiate a
new procurement.
FSS/NFC officials explained that even though NFC experienced
frequent, poor delivery performance on contracts with FPI, it
typically did not seek waivers, because its staff had a large
workload and the time invested attempting to get a waiver would be
unproductive. They added that in many instances, FPI did not respond
to FSS/NFC staff efforts to coordinate on delinquent deliveries. In
these cases, FSS/NFC was left without any remedies other than to wait
until FPI finally shipped the products.
DLA headquarters officials said that they typically did not seek
waivers, because they planned for deliveries far in advance; they had
a good working relationship with FPI that promoted reliable
performance; and their procurement staff visited the factories to
detect and avert delays as they developed. However, these
headquarters officials and the three supply centers said that DLA had
on occasion requested waivers in the past for such reasons as FPI's
inability to produce an item or DLA's desire to purchase part of the
requirement commercially.
Both SSA officials we contacted said that they seldom sought waivers
for timeliness, but SSA/OCG used the waiver process for reasons
related to price. The SSA/OCG official we spoke with estimated that
waivers were requested in 10 percent of SSA's potential purchases
from FPI, mostly for reasons related to price. However, FPI granted
a waiver on one major systems furniture order in 1995, because
arbitrators for SSA's unionized employees required the installation
of systems furniture by a certain date in 1997 that FPI could not
meet.
VA officials said that they had sought waivers from buying FPI
products for reasons related to timeliness, including the
headquarters renovation example mentioned before. However, the VA
officials we contacted said that because their agency is so
decentralized, they could not comment on whether VA typically sought
waivers for reasons related to timeliness. They said, however, that
if they had gotten a waiver for the casegoods, seating, and signage
for the headquarters renovation, they believed they would not have
had nearly as many problems. They also said that the process of
getting waivers was resource intensive. For example, prior to VA's
legal opinion that eyeglasses were not subject to the mandatory
source requirement, they said that VA's medical centers had to
individually develop data and justify waivers when they wanted to
purchase prescription eyeware from commercial sources because of
concerns about FPI timeliness.
--------------------
\8 GAO/GGD-98-50, Mar. 16, 1998.
CONCLUSIONS
------------------------------------------------------------ Letter :6
FPI officials recognize the importance of delivering products and
services on time and have data used to track delivery performance and
an approach for evaluating results, including on-time goals for
fiscal years 1997 and 1998. Although FPI as a whole did not meet its
90 percent on-time delivery goal in any month during 1997, it
improved its delivery performance and production times in that year,
and FPI data showed improvements in 1998. Our statistics, as well as
FPI's evaluations of timeliness, provided insights into delivery
performance and identified areas needing improvement. However, it is
important to recognize that the statistics likely overstate on-time
delivery performance, because they did not account for shipping time
for orders with destination due dates. FPI did not have data or a
workable approach that would allow us to account for shipping time
for such orders. Accounting for shipping time where appropriate
would improve the accuracy of future evaluations of timeliness and
provide a better picture of performance.
Furthermore, over one-half of 109 randomly selected orders showed
that the due dates in FPI's system were later than what customers had
originally requested; and FPI had limited documentation explaining
the differences and whether customers were notified of the reasons
for the changes and approved of the revised due dates. We recognize,
as FPI officials pointed out, that due dates initially requested by
customers may be inconsistent with production capabilities regardless
of whether FPI or a private vendor is the supplier, and a number of
valid reasons can exist for due date changes. However, our work
raised questions about which due dates should be used for measuring
timeliness, especially from the customer's perspective. Better
documentation could have given us, as well as FPI, greater insights
into the reasons for due date changes and whether the due dates in
its system were appropriate for measuring timeliness.
Just as our analysis showed a wide variation in FPI performance by
customer agency, officials within DLA, FSS, SSA, and VA had mixed
views on FPI's performance in meeting delivery due dates, as well as
on FPI delivery practices overall. These views ranged from very
negative to extremely positive. Officials from DLA, FSS, and one of
the two offices we contacted in SSA said that delivery due dates were
driven primarily by FPI production capabilities and not agency
mission needs, and some of the officials expressed concern with this
practice. We recognize that arriving at mutually agreeable due dates
may involve negotiation between the supplier and the customer,
whether the supplier is FPI or a private vendor. In addition, the
views expressed by the agency officials we contacted about FPI's
delivery performance overall, as well as the due date setting
process, may not be representative of the views of all FPI customers.
However, our work showed that (1) certain key agency customers were
clearly dissatisfied with FPI delivery performance and, to some
extent, with its practices with regard to setting due dates; (2)
FPI's delivery performance varied widely by customer agency; and (3)
FPI did not evaluate delivery performance by customer agency. Given
this and FPI's stated commitment to total customer satisfaction,
there appear to be opportunities for FPI to begin evaluating and
monitoring delivery performance by customer agency and to promote
better customer relations by attempting to resolve the specific
concerns key customers have with delivery performance.
RECOMMENDATIONS
------------------------------------------------------------ Letter :7
In order for FPI to have a more accurate and reliable measure of
timeliness for use in evaluating its delivery performance, we
recommend that the Director of BOP direct FPI's Chief Operating
Officer to
-- identify orders with destination due dates and account for
shipping time for these orders when evaluating delivery
performance;
-- develop and implement an approach for documenting the reasons
for due date revisions, whether customers were notified of the
reasons for changes, and whether customers approved of revised
due dates; and
-- appropriately consider due date revisions and whether customers
approved of them in evaluating timeliness.
In light of the concerns raised in this report by some of the top
officials from FPI's major buying agencies and in light of FPI's
stated broad commitment to total customer satisfaction, we also
recommend that the Director of BOP direct FPI officials to contact
these key customers to begin the process of resolving problems and
improving relations. In addition, the Director of BOP should direct
FPI's Chief Operating Officer to begin evaluating and monitoring
delivery performance by customer agency to develop data to use in its
efforts to achieve greater customer satisfaction.
AGENCY COMMENTS
------------------------------------------------------------ Letter :8
In written comments dated June 1, 1998, BOP agreed with the report's
conclusion that delivery performance is improving but problems remain
and said that BOP would implement the report's recommendations. In
commenting on the report, BOP specifically acknowledged that FPI
needs to improve its documentation and customer notification
processes when customer due dates cannot be met or are changed. BOP
also provided its perspective on some of the other issues raised in
the report. First, BOP acknowledged the importance of on-time
delivery and highlighted the actions FPI is taking to improve its
performance and the progress it is making. Second, BOP said that FPI
would like to know how it would compare to some of the larger vendors
in delivery performance and that FPI believes that for complex
projects, issues related to delivery performance are not unique to
FPI. Third, FPI was encouraged that some officials in DOD--FPI's
largest customer--were generally pleased with FPI's delivery
performance. BOP also noted that FPI has informed DOD that FPI will
no longer be the sole provider of "Go to War" items unless DOD
specifically makes such a request. Appendix III contains the full
text of BOP's written comments.
We also obtained oral technical comments from FPI's Chief Operating
Officer and his staff and from program officials in the customer
agencies we contacted on various portions of a draft of this report.
These technical comments have been incorporated in this report as
appropriate.
---------------------------------------------------------- Letter :8.1
As agreed with your offices, unless you publicly announce the
contents of this report earlier, we plan no further distribution
until 30 days from its issue date. At that time, we will send copies
of this report to the Attorney General, the Director of BOP, the
Chief Operating Officer of FPI, the Director of the Office of
Management and Budget, the Administrator for Federal Procurement
Policy, and the heads of the customer agencies we contacted. We will
also make copies available to interested congressional committees, as
well as others on request.
Major contributors to this report are listed in appendix IV. If you
or your staffs have any questions, please contact me on (202)
512-8387.
Bernard L. Ungar
Director, Government Business
Operations Issues
TIMELINESS STATISTICS
=========================================================== Appendix I
Table I.1
FPI Timeliness by Month for Fiscal Year
1996
Percent of orders
shipped on or before
Month order due due date
---------------------------------------------- ----------------------
October 1995 68
November 1995 59
December 1995 61
January 1996 66
February 1996 74
March 1996 81
April 1996 81
May 1996 78
June 1996 77
July 1996 75
August 1996 73
September 1996 74
Timeliness for fiscal year 1996 72
----------------------------------------------------------------------
Note: These statistics do not reflect shipping time for those orders
that had destination due dates. In addition, these statistics are
based on due dates from the customer order entry database. The
limitations associated with these factors are discussed in the
report.
Source: GAO analysis of Federal Prison Industries data.
Table I.2
FPI Timeliness by Month for Fiscal Year
1997
Percent of orders
shipped on or before
Month order due due date
---------------------------------------------- ----------------------
October 1996 78
November 1996 72
December 1996 65
January 1997 62
February 1997 67
March 1997 72
April 1997 77
May 1997 79
June 1997 81
July 1997 87
August 1997 84
September 1997 82
Timeliness for fiscal year 1997 76
----------------------------------------------------------------------
Note: These statistics do not reflect shipping time for those orders
that had destination due dates. In addition, these statistics are
based on due dates from the customer order entry database. The
limitations associated with these factors are discussed in the
report.
Source: GAO analysis of Federal Prison Industries data.
Table I.3
FPI Timeliness by Product Category,
Fiscal Years 1996 and 1997
1996 1997
------------------------------ ------------------------------
Percent of Percent of
Product orders shipped Median orders shipped Median
category/ on or before production on or before production
examples due date time (days) due date time (days)
---------------- -------------- -------------- -------------- --------------
Data graphics 75 23 76 14
(signs,
printing,
optics)
Electronics 74 37 73 16
(cables,
connectors,
circuit
assemblies)
Furniture 66 102 78 89
(desks,
bookcases,
ergonomic
chairs)
Metals 86 63 75 66
(shelving,
lockers,
casters)
Systems 75 97 85 78
furniture
Textiles 77 52 83 49
(clothing,
linens,
mailbags)
--------------------------------------------------------------------------------
Note 1: These statistics do not reflect shipping time for those
orders that had destination due dates. In addition, these statistics
are based on due dates from the customer order entry database. The
limitations associated with these factors are discussed in the
report.
Note 2: Median production times exclude quickship items, which FPI
aims to ship within 30 days. Median production times for quickship
items were 21 days in both 1996 and 1997.
Source: GAO analysis of Federal Prison Industries data.
Table I.4
FPI Timeliness by Customer Agency,
Fiscal Years 1996 and 1997
Percent of Percent of
orders shipped orders shipped
on or before due on or before due
date in fiscal date in fiscal
Agency year 1996 year 1997
---------------------------------- ---------------- ----------------
Agriculture 74 80
Air Force 70 77
Army 72 75
Commerce 76 78
Defense Logistics Agency 66 71
Department of Defense 69 77
(miscellaneous nonservice
components)
Energy 77 83
Environmental Protection Agency 54 73
General Services Administration 77 74
Health and Human Services\a 60 73
Housing and Urban Development 76 69
Interior 71 80
Judiciary 88 92
Justice 78 79
Labor 86 86
Marine Corps 71 73
Navy 69 73
Postal Service 86 91
State 55 70
Transportation 75 78
Treasury 72 72
Veterans Affairs 47 47
----------------------------------------------------------------------
Note: These statistics do not reflect shipping time for those orders
that had destination due dates. In addition, these statistics are
based on due dates from the customer order entry database. The
limitations associated with these factors are discussed in the
report.
\a FPI's system coded Social Security Administration (SSA) orders as
Health and Human Services (HHS) orders; SSA became an independent
agency in the executive branch in 1995. SSA orders accounted for
over one-half of the orders in fiscal year 1997 that were coded as
HHS.
Source: GAO analysis of FPI data.
OBJECTIVES, SCOPE, AND METHODOLOGY
========================================================== Appendix II
Our objectives were to (1) develop and assess statistics on FPI
delivery performance and (2) obtain the views of selected customer
agencies' procurement officials on FPI delivery practices. As part
of this work, we considered FPI's efforts to track delivery
performance and its practices with regard to setting due dates. Our
work for the second objective primarily involved holding discussions
with selected customer agency officials about their experiences with
FPI delivery performance, the due date setting process, their efforts
to monitor FPI delivery performance, and available remedies for late
deliveries.
To meet our first objective, we obtained and analyzed computerized
data involving over 140,000 agency orders for over 600,000 line items
of products and services from FPI's customer order entry database
(COED) for fiscal years 1996 and 1997.\9 FPI uses COED, which is
maintained at its Customer Service Center in Lexington, KY, to track
and record information on agency orders for products and services.
Agency orders can contain one to several hundred line items. Each
line item identifies a specific product and quantity and has a
distinct due date. Large orders with many line items can have
multiple products and due dates. Because orders vary tremendously, a
key factor in determining whether orders are delivered on time is how
an order is defined. For its own evaluations of timeliness, FPI
groups line items within orders by due date and factory. FPI defines
each of these groupings--which essentially are the shipments expected
for an order from a given factory on a given date--as individual
orders for the purpose of measuring timeliness. This approach allows
FPI to track the performance of individual factories on a weekly and
monthly basis.
For our analysis, we generally followed the approach FPI uses to
define orders and developed timeliness statistics by month, year,
product category, and customer agency. Specifically, we grouped the
line items within each order by date and compared the shipping date
for each line item with its due date, as FPI does for its timeliness
indicator. If one of the line items within each grouping was late,
that grouping, or order, was counted as late. This approach allowed
us to develop monthly data that we could use to examine trends. We
also developed data showing FPI's delivery performance for each of
its major customer agencies using this approach. To identify these
agencies, we analyzed FPI's fiscal year 1996 sales report.
For our analysis by month, year, and customer agency, we did not
group the line items by factory as FPI does because we did not focus
on the performance of individual factories. However, we had to use
this additional grouping when isolating orders from factories that
produce products and services in six product categories--data
graphics, electronics, furniture, metals, systems furniture, and
textiles. We also developed timeliness statistics for products in
FPI's quickship program. FPI guarantees that it will ship products
in this program within 30 days of receipt of agencies' orders.
Products in FPI's quickship catalogue include certain types of
ergonomic chairs and other furniture, linens, clothing, targets, and
traffic signs.
As part of our work, we also obtained and analyzed the results of
FPI's internal timeliness measure for fiscal year 1997 and the first
half of 1998. We compared FPI's results with our fiscal year 1997
analysis and reconciled any differences. We did not independently
verify FPI's analysis for fiscal year 1998. We also examined how FPI
accounts for shipping time for those orders where the due date is
when deliveries should arrive at the customer's location, not the
date by which they should be shipped. Due dates specified by
customers can be origin--meaning that the orders should be shipped by
the due date--or destination, meaning that the orders should arrive
at the customer's location by the due date. In doing our work, we
did not determine the reasons individual agency orders were delivered
late or assess the effect of individual late deliveries on federal
agencies. In addition, we rounded the timeliness statistics
presented in this report to the nearest percentage.
We performed a limited reliability and validity check of FPI's COED
data by comparing a sample of hard copy customer orders with the data
in the COED system. According to an FPI official knowledgeable of
customer orders, FPI received more than 60 percent of its orders by
hard copy in fiscal year 1997. The remainder of the orders were
submitted electronically. We did not attempt to validate any
electronic orders because printouts with data from these orders that
FPI showed us had much less information than the hard copy orders.
We selected a random sample of 240 customer orders from orders that
were placed in fiscal year 1997. This sample was designed to ensure
that we selected orders that had many line items as well as orders
that had only a few line items. In addition, we oversampled because
we knew that many of the orders probably would have been placed
electronically and would not have hard copies.
FPI sent us copies of the original customer orders and related
documentation that were available, such as correspondence to and from
customers, for 136 of the orders. Five of FPI's six major product
categories were represented in these orders. Only electronics, which
accounted for less than 7 percent of orders in fiscal year 1997, was
not represented. The remaining orders from the 240 we randomly
selected were unavailable because either the agencies had filed their
orders electronically, or the orders were not centrally located and
could not be obtained in a timely manner from the individual
factories. FPI officials told us that they sent us all the
documentation that was readily available for the 136 orders. In
addition, FPI provided us with access to its COED database in
Lexington, which enabled us to review all the electronic information
that was available for these orders. We reviewed the information
that was available for these 136 orders and determined that 109 had
due dates that customers had specified.
Our timeliness analysis used three dates from the COED system. These
were the dates the orders were entered into the COED system (order
date), the dates the orders were due (due date), and the dates the
orders were shipped (last ship date). We could not focus on the
order dates because we knew there would be a time lag between the
date the order was signed and the date it was entered into the COED
system. In general, we noticed that this time lag was between 2 to 3
weeks. However, we could not assess whether these time lags were due
to (1) delays by the customer in mailing the order, (2) the
timeliness of the mail carrier, or (3) delays by FPI in entering the
orders into the COED system. We could not assess the reliability of
the last ship dates, because we did not have the resources to contact
the agencies directly and find out if they had records of receiving
the items FPI had shipped to them. Therefore, we focused on the due
dates agencies had specified in the orders.
We compared the customer-requested due dates with the due dates
entered into COED. We attempted to document any instances where a
due date for a line item in COED was different from the due date the
agency had originally requested in its order. Using the information
available for these orders, we also attempted to document the reasons
for due date changes and other pertinent information, such as whether
customers were notified of the reasons for due date changes and
whether they approved of the revisions. We also documented whether
customers had specified that the due dates in the orders were for
destination or origin. For orders where we could not determine the
reason for due date changes, FPI staff at the Customer Service Center
in Lexington researched the orders to identify the reason for date
changes, as well as other information, such as whether the customer
was notified regarding any due date changes that may have occurred
and approved of them. In addition, FPI headquarters officials
provided us with other documentation for these orders at the end of
our review. The results of our work related to the 109 orders are
not projectable to the universe of agency orders placed in 1997
because we were only able to obtain usable information on 109 of the
240 orders in our original sample, which represented a response rate
of about 45 percent. This response rate was mainly due to the fact
that many of the orders were placed electronically rather than
through hard copy.
To meet the second objective, we interviewed key procurement
officials from four agencies that we judgmentally selected who were
among FPI's largest buyers: the General Services Administration's
(GSA) Federal Supply Service (FSS) in Arlington, VA; the Defense
Logistics Agency (DLA) in Ft. Belvoir, VA; the Social Security
Administration (SSA) in Baltimore, MD; and the Department of Veterans
Affairs (VA) in Washington, D.C. Collectively, these agencies
accounted for over one-quarter of FPI's 1996 sales of $496 million.
DLA was the second largest buying component within DOD, next to the
Army. The other three agencies were in the top four buyers among
civilian agencies. In addition to contacting headquarters
procurement officials at these agencies, we also contacted one of
GSA's nationwide commodity centers, FSS' National Furniture Center
(NFC), because furniture is FPI's largest product line. We contacted
three DLA supply centers: the Defense Supply Center Philadelphia
(DSCP) in Philadelphia, PA; the Defense Supply Center Columbus (DSCC)
in Columbus, OH; and the Defense Supply Center Richmond (DSCR) in
Richmond, VA. Within SSA, we spoke with officials from the two
buying components that purchase products from FPI, the Office of
Property Management (SSA/PM) and the Office of Operations, Contracts,
and Grants (SSA/OCG). Within VA, we spoke with officials from the
Office of Administration and the Office of Acquisition and Materiel
Management, as well as officials in the Veterans Health
Administration (VHA) from the offices of Patient Care Services and
Environmental Management Services.
We obtained information and views from these officials on several
topics related to timeliness, including how FPI performed in meeting
due dates; how specific due dates were set; whether due dates were
driven by FPI production capabilities or mission needs; what
processes were used to monitor FPI's delivery performance and that of
private vendors; and what remedies were available in the event of a
late delivery from FPI and private vendors, including a discussion of
FPI's waiver process. Although the views officials at these agencies
expressed and the information they provided may not be representative
of all FPI customers, they provided useful insights into FPI delivery
practices from the customer's perspective. We did not verify data we
obtained related to our discussions with agency officials. Related
to our discussion with these officials about the FPI waiver process,
we used FPI's waiver database to determine the extent to which FPI
granted customers waivers from buying FPI products for reasons
related to timeliness.
We did our work between July 1997 and May 1998 in accordance with
generally accepted government auditing standards. We requested
written comments on this report from the Director, Bureau of Prisons
(BOP). These comments are discussed near the end of the letter and
are reprinted in appendix III. FPI also provided oral technical
comments, which we considered in preparing the final report. We also
held exit conferences with program officials of the customer agencies
we visited to verify applicable data, facts, and views presented in
this report.
(See figure in printed edition.)Appendix III
--------------------
\9 Because of concerns we had about the manageability of such a large
volume of data, we did not request all the variables from COED and
did not request data on blanket orders, which make up about 3 percent
of the orders.
COMMENTS FROM THE BUREAU OF
PRISONS
========================================================== Appendix II
(See figure in printed edition.)
(See figure in printed edition.)
(See figure in printed edition.)
The following are GAO's comments on the Bureau of Prisons' letter
dated June 1, 1998.
GAO COMMENTS
1. BOP said that it is not FPI's policy that due dates be driven by
production capabilities rather than agency missions. This seemed
different from the statements made by FPI's Chief Operating Officer
during the review that it often did not accept the due dates
customers requested because they were inconsistent with FPI's
production capabilities. Seeking clarity, we contacted an FPI
official who was involved in preparing the written comments. This
official said that the intent of the comment and the paragraph that
followed was to emphasize an important point. That is, if FPI does
not accept an agency-requested due date and proposes alternative
dates that would impede the customer agency's mission, FPI would
likely issue a waiver if the customer requested one.
2. BOP pointed out that the random sample of orders that we used to
test the accuracy and reliability of the due dates that FPI used to
measure timeliness was not representative of the orders FPI receives.
Specifically, BOP was concerned that a disproportionate number of
high-dollar value systems furniture orders led to a skewing of the
information collected because these orders have more frequent due
date changes. This condition however, does not affect our
conclusions on this matter or the overall message of the report. As
discussed in the report, we recognize that the specific results from
our work related to the 109 orders cannot be projected to the
universe of FPI orders. However, the results of our analysis did, as
FPI agreed in its written comments, raise questions about the due
dates being used to measure timeliness, especially from the
customer's perspective.
3. BOP said that of the 61 orders that involved due dates that were
later than the customer originally requested, 23 did not involve a
change. According to BOP, the delivery dates requested by customer
agencies for these orders did not provide sufficient lead time and,
therefore, FPI had to revise the dates as a condition of accepting
the order. BOP acknowledged in its written comments that
nevertheless, FPI needs to improve its documentation and customer
notification processes when customer due dates cannot be met or are
changed. As we discussed in the report, the lack of documentation
showing the reasons due dates in FPI's system were different from
what customers requested and whether customers were notified of the
reasons and approved of the changes impeded our ability to assess the
appropriateness of many of the due dates for measuring timeliness.
MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix IV
GENERAL GOVERNMENT DIVISION,
WASHINGTON, D.C.
Gerald Stankosky, Assistant Director
David E. Sausville, Evaluator-in-Charge
Martin de Alteriis, Senior Social Science Analyst
Joanne D. Meikle, Senior Social Science Analyst
Delois Richardson, Computer Specialist
Catherine M. Hurley, Senior Computer Specialist
William J. Dowdal, Senior Evaluator
DALLAS FIELD OFFICE
Robert T. Griffis, Senior Evaluator
Patricia Sari-Spear, Senior Evaluator
*** End of document. ***