Social Security Administration: Additional Actions Needed in	 
Ongoing Efforts to Improve 800-Number Service (08-AUG-05,	 
GAO-05-735).							 
                                                                 
The Social Security Administration (SSA) at some point touches	 
the life of nearly every American. Each day thousands of people  
contact SSA to file claims, update records, and request 	 
information from its 1,300 field offices, website, and national  
toll-free 800 number. Implemented nationwide in 1989, SSA's	 
800-number has become a principal contact point for millions of  
individuals seeking agency services. Congressional requesters	 
asked GAO to review the quality of SSA's 800 number in terms of  
caller access and agent accuracy of response and courtesy.	 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-05-735 					        
    ACCNO:   A32278						        
  TITLE:     Social Security Administration: Additional Actions Needed
in Ongoing Efforts to Improve 800-Number Service		 
     DATE:   08/08/2005 
  SUBJECT:   Customer service					 
	     Data integrity					 
	     Employee training					 
	     Federal social security programs			 
	     Government information dissemination		 
	     Internal controls					 
	     Telephone services 				 
	     Training utilization				 
	     Performance measures				 
	     Program evaluation 				 
	     Policies and procedures				 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO 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.                                         **
**                                                              **
******************************************************************
GAO-05-735

United States Government Accountability Office

GAO

                Report to the Committee on Finance, U. S. Senate

August 2005

                                SOCIAL SECURITY
                                 ADMINISTRATION

Additional Actions Needed in Ongoing Efforts to Improve 800-Number Service

GAO-05-735

[IMG]

August 2005

SOCIAL SECURITY ADMINISTRATION

Additional Actions Needed in Ongoing Efforts to Improve 800-Number Service

  What GAO Found

Despite making improvements to its 800-number service, SSA still has
difficulty keeping pace with caller demand for agent assistance. In 2001,
SSA upgraded its 800-number network so that all callers could either
access its automated services or be routed to the next available agent at
any site in the network-a feat not possible under the previous system. The
new network also enhanced SSA's ability to monitor and manage call
traffic, agent availability, and network operations in real-time to ensure
the network's integrity and the consistent delivery of services. SSA also
expanded its automated and agent-assisted services accessible through the
800-number network. However, SSA's expansion of its automated services to
reduce agent call burden has not had its intended effect, as callers
continue to show a strong preference for agent assistance. In fiscal year
2004, about 51 million of the more than 71 million callers requested to
speak to an agent. However, 8.7 million, or 17 percent, of these calls did
not get through to an agent-a 2 percentage point increase over the
previous year.

SSA has taken steps to help agents provide callers with accurate
information and consistent services, but still has problems with agents
assisting callers in line with agency policies and procedures. SSA's
training curriculum provides agents with a comprehensive overview of SSA
programs. Agents are also encouraged to use available on-the-job
resources, including a customized computer application that helps agents
provide consistent service and accurate responses. Nevertheless, from 2001
through 2003, SSA did not meet its 90 percent target for service
accuracy-that is, agents' performance in handling non-payment related
issues in accordance with agency requirements. Although SSA has taken
several actions to help agents improve their performance, including
mandating agent use of the computer application, it has not yet determined
why agent compliance with agency policies continues to fall short.

SSA trains and monitors agents for courtesy and conducts periodic customer
satisfaction surveys, but does not routinely capture all customer
complaints about alleged agent discourtesy. Agents receive training on
developing their interviewing and interpersonal skills, and SSA monitors
agents to determine whether or not they are providing courteous service to
callers. SSA monitoring indicates that agent courtesy levels are high. SSA
solicits limited customer feedback on agent courtesy in its annual surveys
and compiles general ratings, but these surveys do not ask callers for the
reasons behind the ratings. Callers to the 800 number do complain of agent
discourtesy, but SSA does not routinely document and assess all
complaints. Some call center staff told us that when they receive
allegations of agent discourtesy, they typically apologize for the
discourteous service and may proceed to assist the caller without
recording the complaint. SSA has feedback mechanisms in place to capture
caller complaints, but these mechanisms do not do so in a manner that
allows SSA to assess complaints and identify corrective actions needed.

United States Government Accountability Office

Contents

  Letter

Results in Brief
Background
SSA Has Improved Overall Access to the 800 Number, but Many

Calls Seeking Agent Assistance Do Not Get Through SSA Trains and Provides
Agents On-the-Job Resources, but Agents Have Not Met SSA's Standard for
Accuracy of Assistance SSA Conducts Training, Monitoring, and Customer
Surveys but

Lacks a Uniform System for Assessing Agent Courtesy Conclusions
Recommendations for Executive Action Agency Comments and Our Evaluation

                                       1

                                      2 4

                                       9

18

27 31 32 32

Appendix I Objectives, Scope, and Methodology

Appendix II Comments from the Social Security Administration

Appendix III GAO Contact and Staff Acknowledgments

  Tables

Table 1: Major Services That Agents Provide to 800-Number Callers 9 Table
2: Services Available through the 800-Number Automated Menus, As of June
2005 13 Table 3: Top Five Types of Service Errors Observed in Fiscal Year
2003 25

  Figures

Figure 1: SSA's 800-Number Call Center Locations 6 Figure 2: Most Frequent
Reasons for Calls to the 800 Number, Fiscal Year 2003 8 Figure 3: Calls
Placed to the 800-Number, Fiscal Years 2002 through 2004 (in Millions) 15
Figure 4: 800-Number Calls Seeking Agent Assistance That Did Not Get
Through, Fiscal Years 2002 through 2004 (in Millions) 16

Figure 5: 800-Number Calls Handled and Abandoned in

Automation, Fiscal Years 2002 through 2004 (in Millions) 17 Figure 6:
SSA's Performance in Meeting Payment Accuracy and

Service Accuracy Targets, Fiscal Years 1998 through 2003 23 Figure 7:
Regional Performance for Meeting Service Accuracy

Target, Fiscal Years 2001 through 2003 24

Abbreviations

CHIP Customer Help Information Program
OIG Office of the Inspector General
OQA Office of Quality Assurance and Performance Assessment
OTS Office of Telephone Services
SSA Social Security Administration
SSI Supplemental Security Income

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States Government Accountability Office Washington, DC 20548

August 8, 2005

The Honorable Charles E. Grassley
Chairman
The Honorable Max Baucus
Ranking Member
Committee on Finance
United States Senate

The Social Security Administration (SSA) at some point touches the life of
nearly every American. Its services include issuing Social Security cards,
maintaining earnings records, and administering some of the largest
federal benefit programs-Old Age and Survivors Insurance and Disability
Insurance-commonly referred to as Social Security-and Supplemental
Security Income (SSI). Each day thousands of people contact SSA's 1,300
field offices, Web site, and national toll-free 800 number
(1-800-SSA-1213)
to file claims, report events to update records, and request information
about SSA's programs. Since becoming available nationwide in 1989, SSA's
800 number has become a principal contact point for individuals seeking
agency services. Callers to the 800 number may opt to use a self-service
menu of automated services or request to speak with an agent. In fiscal
year 2004, callers placed over 71 million calls to SSA's 800 number.

SSA has made improvements and adjustments to its 800-number service in
response to customer feedback. However, some problems have persisted.
For example, SSA reported that millions of calls were encountering busy
signals or being terminated before being served. An SSA survey of 800
number callers in September 2000 found that callers wanted easier and
faster access to live service-that is, to speak with an agent without
wading through the lengthy list of automated services. Furthermore,
callers commented that SSA could improve its 800-number services by
having agents with expertise provide consistent information and exhibit
courteous behavior and helpful attitudes.

In this context, you asked us to review the quality of SSA's 800 number in
terms of caller access, agent accuracy of response and courtesy. This
report discusses SSA's efforts to (1) improve caller access to the 800
number; (2) ensure that SSA agents provide accurate responses and follow
SSA requirements; and (3) ensure that agents treat callers courteously.

  Results in Brief

To conduct our work, we reviewed SSA's policies and procedures for
800-number agents and interviewed SSA officials to develop information on
SSA's telephone systems and services, training offerings and requirements,
and the steps they take to ensure that their agents provide acceptable
service. Specifically, we interviewed and obtained documentation from
officials at SSA headquarters in Baltimore, Maryland; regional officials
in 3 of the 10 regions; and staff working in a total of six call
centers-two in each location visited-Birmingham, Alabama; Kansas City,
Missouri; and Richmond, California. The call centers we selected varied in
the volume of calls handled and included those that handled 800-number
calls on a routine basis and centers that handled calls as needed. We
reviewed SSA performance data related to access, accuracy, and courtesy
and determined the data to be sufficiently reliable for the purposes of
this report. We conducted our work from September 2004 through July 2005
in accordance with generally accepted government auditing standards.
Appendix 1 provides a more detailed description of our scope and
methodology.

Despite making improvements to its 800-number systems, management, and
services to improve caller access, SSA still has difficulty keeping pace
with caller demand for agent assistance. In 2001, SSA upgraded its
800-number network so that all callers could either access its automated
services or be routed to the next available agent at any site in the
network-a feat not possible under the previous system. The new network
also enhanced SSA's ability to monitor and manage call traffic, agent
availability, and network operations in real-time to ensure the network's
integrity and the consistent delivery of services. SSA also expanded its
automated and agent-assisted services accessible through the 800-number
network. However, SSA's expansion of its automated services to reduce
agent call burden has not had its intended effect, as callers continue to
show a strong preference for agent assistance. In fiscal year 2004, 51
million of the more than 71 million callers opted to speak with an agent,
but 8.7 million, or 17 percent, did not get through-a 2 percent increase
over the previous year.

SSA has taken steps to help agents provide callers accurate information
and to comply with SSA requirements, but still has problems with agents
not meeting its standards for accurate service. SSA provides agents with
comprehensive training and equips them with on-the-job resources such as a
customized computer application to enable them to offer a broad range of
services and to help ensure that they provide accurate and consistent
service. SSA also makes experienced staff available to assist agents with

complex or technical calls. In addition, SSA monitors agents' calls
agencywide to assess agent accuracy and training needs. SSA's own
monitoring assessments for 1998 through 2003 found that the agency
generally met its 95 percent standard for "payment accuracy," having
agents correctly handle inquiries involving eligibility and benefit
payment issues. However for this same period, SSA did not meet its
standard for "service accuracy," having agents correctly handle issues
that do not have the potential to affect eligibility and/or payment
according to agency requirements. SSA's overall performance for "service
accuracy" for fiscal year 2003 was 85.1 percent; below the agency's 90
percent target. According to SSA's assessment, agents' failure to obtain
six personal identifying pieces of information from callers to verify
their identity before accessing and disclosing information was the most
frequent service error committed by agents. This error accounted for 28
percent of all service errors identified through SSA's assessment in
fiscal year 2003. SSA has taken several actions to help agents improve
their performance, including requiring agents to use the computer
application designed to help them comply with agency requirements by
directing their questions and responses to callers. However, the
improvement in the service accuracy rate that followed these initiatives
was short-lived. SSA has not determined why agents fail to comply with
requirements and thus commit service errors.

SSA trains and monitors agents for courtesy and conducts periodic customer
satisfaction surveys, but does not routinely document or analyze all
incidents of discourtesy or caller complaints. As part of the basic
training curriculum, agents are taught interviewing and interpersonal
skills. In addition to monitoring for accuracy, SSA also monitors agents
to determine whether or not they were courteous to callers. Based on 4,384
monitored calls in 2003, SSA reported an agent courtesy rate of 99.9
percent. Although this would mean discourtesy was highly infrequent, it
would still mean that agents may have handled as many as 60,000 calls
placed during fiscal year 2003 discourteously. SSA solicits direct
customer feedback about agent courtesy by surveying callers annually. In
2004, 91 percent of the respondents rated the agent level of courtesy as
good, very good, or excellent while 5 percent rated it as fair and 4
percent rated it as poor or very poor. However, the survey does not ask
callers their reasons for the ratings. Also, while callers also contact
SSA to complain of agent discourtesy, SSA does not routinely document all
complaints. We were told that when callers report allegations of agent
discourtesy, call center staff typically apologize for the discourteous
service and may proceed to assist the caller without making a record of
the complaint. In addition, callers sometimes make complaints about agent
discourtesy

through SSA's Web site. However, the Web site does not give guidance to
complainants about the specific information they should provide. As a
result, customers do not typically provide information that would help SSA
assess the overall nature of complaints and identify any needed action.
According to experts on customer service, a good management complaint
system is key to building customer relations and identifying recurring
problems and solutions to prevent future problems. Without such a system,
SSA may be missing opportunities to address customer concerns and improve
800-number services.

This report contains recommendations to the SSA Commissioner that are
intended to improve the quality of the 800-number service related to
access, accuracy, and courtesy. In its comments on a draft of this report,
SSA agreed with our recommendations to identify cost-effective ways to
increase agent availability to handle 800-number calls and conduct a
comprehensive analysis of the source of agent service errors. SSA
disagreed with our recommendation to establish procedures for documenting
and assessing customer complaints. SSA believes a formal complaint system
is not necessary or cost-effective given the consistently high level of
courteous service indicated by SSA's service monitoring and customer
satisfaction surveys. In our report, we acknowledge the high level of
agent courtesy indicated through these sources. Yet, SSA's high call
volume means that even with low rates of discourtesy, agents may be
treating potentially tens of thousands of callers discourteously. Good
customer service stresses the importance of paying attention to customers'
complaints, however minor, and establishing a simple userfriendly and
comprehensive complaint management system. We do not anticipate a need for
SSA to expend extensive resources in meeting our recommendation, but
rather suggest that SSA modify and uniformly apply the mechanisms already
in place to obtain information about callers' concerns or complaints that
could help the agency further improve customer service.

Background 	Before implementing the nationwide 800-number service, SSA
delivered most of its services to the public face-to-face in an SSA field
office. In 1989, SSA implemented a national, toll-free 800 number to
better enable individuals to request information on SSA programs or report
events that affect their own or someone else's SSA records or payments.1
SSA set up

1First established in 1988, SSA's 800-number service was initially
available to 60 percent of the population. SSA expanded the service to all
callers nationwide in 1989.

the 800-number service with the expectation that callers would ask basic
questions and conduct simple business transactions, such as reporting
address changes and scheduling field office appointments. When a call came
into the 800 number, it would be routed to a local SSA call center. This
strategy resulted in a high busy rate. Troubled by high busy-signal rates,
SSA in 1996 added a nationwide automated menu to the 800 number that
allowed callers to conduct a limited number of transactions without
speaking to an agent.

In 1997, we identified a number of conditions that limited the
effectiveness of SSA's 800-number service.2 For one, callers often reached
a busy signal instead of the automated menu or an agent. In addition, the
automated menu offered only a limited number of services. To reach an
agent, callers were required to select a specific topic about which they
wished to speak to an agent so that the system could direct their call to
an agent in a call center with the requisite subject matter expertise.
This routing strategy led to some call centers being overwhelmed with
calls. Also, because agents could not transfer calls, callers sometimes
were inconvenienced by having to redial the 800 number to complete their
business.

Since the introduction of its nationwide 800-number service, SSA has
worked to keep pace with the public's growing demand for telephone
services and interests in conducting more complex transactions over the
telephone. Today, calls made to the 800 number are answered at 44
geographically dispersed locations. A call placed to the 800 number may be
answered by agents located in any one of SSA's 36 teleservice centers, 6
Program Service Centers; or at one of 2 components within SSA's Office of
Central Operations. 3 Figure 1 shows the locations of these call centers
within the 10 SSA regions.

2GAO, Social Security Administration: More Cost-Effective Approaches Exist
to Further Improve 800-Number Service, GAO/HEHS-97-79 (Washington, D.C.:
June 1997).

3Two components within SSA's Office of Central Operations also answer
800-number calls. SSA's Wilkes Barre Data Operations Center in
Pennsylvania has approximately 266 agents trained to answer calls. In
addition, SSA's Division of Earnings Record Operations in Baltimore,
Maryland, has 198 agents trained to answer 800-number calls.

Figure 1: SSA's 800-Number Call Center Locations

Source: GAO analysis of SSA data.

SSA staffs its 36 teleservice centers with approximately 4,060 teleservice
representatives who answer incoming calls to the 800 number. In addition,
each of SSA's six program service centers, which are co-located on
teleservice center sites, has designated specialists, called "SPIKES," who
have been cross-trained to provide back-up support in answering 800-number
calls during peak call volume periods. The SPIKE staff is comprised of
various technical staff in the program service centers whose routine
responsibilities include processing claims, mailing out notices,

managing SSA's debt collection activities, and handling reports of
nonreceipt of checks and representative payee issues.

SSA employs a cadre of approximately 2,030 trained SPIKES in its six
program service centers. When the volume of calls is expected to exceed
the levels that teleservice representatives can handle, SSA activates
SPIKES, diverting them from their routine responsibilities to answer
incoming 800-number calls. These peak calling periods typically occur on
the first day of the week, the first week of the month, and the first 3
months of the year. In this report, we refer to teleservice
representatives and SPIKES as "agents" and to teleservice centers and
program service centers as "call centers."

SSA's Office of Telephone Services (OTS) plans, implements, operates, and
evaluates SSA telephone service to the public delivered by way of the
national 800 number and field offices. OTS plans and conducts studies,
pilots, and analyses of 800-number and field office telephone operations
to assess and improve the service. It also provides direct support to call
centers and field offices, including developing and communicating uniform
operating policies and procedures. OTS staff works closely with SSA's
vendor that supplies and manages the network hardware, software, and
telephone equipment used to support the 800-number service. OTS also
manages the 800-number network operations, designs and administers call
routing plans, and monitors call handling and adjusts call routing to
handle emergency situations.

Full-time SSA agents spend much of their time answering calls.4 These
calls may cover a broad range of inquiries about SSA programs and
procedures. Figure 2 shows the 10 most frequent reasons for calls to the
800 number in fiscal year 2003.

4The terms full-time and back-up agents are used to describe SSA employees
who handle 800-number telephone calls as their primary and secondary
responsibility, respectively.

Figure 2: Most Frequent Reasons for Calls to the 800 Number, Fiscal Year
2003

Reason

File a claim 10.7

Social Security Number -related issues

Change of address (other than Supplemental Security Income)

Direct deposit

Check inquiry

Entitlement/eligibility

Claims status inquiry

Benefit verification

Medicare coverage

Nonreceipt of payment

0 2 4 6 8 101214161820

Percentage of total calls

Source: SSA weighted estimates based on sampled data.

Agents' time off the phone, such as for staff meetings, training sessions,
or annual leave must be scheduled months in advance so that the network
operations may continue without interruption.

SSA sets goals for telephone access and agent services and measures
performance in these areas. In recent years, to measure access, SSA
calculated the number of calls handled, the number of calls that reach the
800 number on their first attempt, and the number of calls that reach an
agent within 5 minutes of selecting the option to speak with an agent. In
fiscal year 2005, SSA replaced these measures with two new access
performance measures-the average speed of answer and the agent busy
rate-consistent with standards in the telecommunications industry. SSA
also expects agents to adhere to agency guidance and procedures and sets
standards and measures agent accuracy (i.e., compliance with SSA's
requirements when serving callers) and agent courtesy. The Office of

Quality Assurance and Performance Assessment (OQA) measures the accuracy
of information agents provide callers by listening in daily to a
statistical random sample of calls handled by agents nationwide. OQA
assesses accuracy based on whether agents adhered to SSA requirements when
responding to callers' inquiries. As shown in table 1 agents are expected
to provide callers a broad range of services. OQA also periodically
surveys 800-number callers to assess, among other things, callers'
perception of agent courtesy.

Table 1: Major Services That Agents Provide to 800-Number Callers

Provide information about Social Security numbers including requirements
for obtaining Social Security numbers and replacement cards

Explain SSA program rules specific to individuals' circumstances

Screen individuals for entitlement to benefits under SSA programs

Advise individuals about their rights and responsibilities as participants
in SSA programs

Receive and input reports such as address changes, non-receipt of checks,
and deaths that affect SSA records or benefit payments

Answer questions about communications individuals received from SSA

Inform individuals of assistance provided by other agencies and make
referrals

                       Source: GAO analysis of SSA data.

  SSA Has Improved Overall Access to the 800 Number, but Many Calls Seeking
  Agent Assistance Do Not Get Through

Despite making improvements to its 800-number systems, SSA still has
difficulty keeping pace with caller demand for agent assistance. Since
2001, SSA has made improvements to its telephone systems, management, and
services to improve caller access to the 800-number network. Specifically,
the new enterprise-wide network improved incoming call routing and network
capacity; enhanced SSA's ability to manage network operations, forecast
call volumes, and set staffing levels; and expanded automated and agent
services. However, callers continue to demonstrate a preference for
speaking with an agent over using the automated service menus. In fiscal
year 2004, about 51 million callers requested to speak to an agent. Of
these calls, 8.7 million, or 17 percent, of these calls did not get
through to an agent-a 2 percent increase over the previous year.

SSA Made Numerous SSA upgraded the network to help overcome past access
problems. One Upgrades to Its 800-major upgrade was the replacement of the
geographically based routing Number Network system with a nationwide
routing system capable of distributing calls to

any agent within the network. This change gave SSA the ability to monitor
call traffic and agent availability in real time at each call center and
receive "cradle to grave" management information on a call's movement from
the

time the caller dials the 800 number until the call is terminated. The
network also effectively eliminated the busy signal that callers
encountered when using the older system. The new system accepts all calls
made to the 800-number network and provides callers with a broad range of
automated services. Calls seeking agent assistance are distributed to 1 of
SSA's 44 answering sites.

When callers dial the 800 number, the network provides a series of prompts
to direct them to the desired services. The network uses recorded
announcements and pre-set menu prompts to separate callers according to
language preference (i.e., English or Spanish) or type of telephone
service (i.e., touchtone or rotary dial). The network uses a digitized
voice to read menu selections to the caller and responds to caller-entered
touch-tone digits. The caller's selection can invoke a number of options,
such as playing a recorded announcement (e.g., on cost-of-living
adjustments) or transferring a call to an agent. SSA provides callers with
an extensive menu of available automated services before offering them the
option of acquiring agent assistance. SSA told us that the menus were set
up this way to offer callers an opportunity to conduct their business
using automated services before forwarding their calls to agents.

When a caller indicates a preference for agent assistance, the network
determines the optimum destination for the call. It reviews among other
factors agent availability, the number of calls in queue, and the minimum
expected delay. If all agents are busy and call queues are filled to
capacity, the network delivers an agent busy message to callers, advising
them that heavy call volume prohibits the transfer of their call to an
agent and encouraging them to call back during periods of typically lower
call volumes.

A call placed in agent queue remains queued until an agent becomes
available. The network applies treatments to calls waiting in agent queue,
such as announcements promoting the use of SSA's Web site. According to
SSA, if the wait time in an agent's queue exceeds 15 minutes, the call is
rerouted to another agent and given priority over other incoming calls.
The network continually tracks the status of each call until the caller
disconnects. Although the network was designed to hold one call per agent
in queue, the vendor told us that it typically holds up to 1.65 calls in
queue per agent.

SSA and the vendor have taken measures to ensure the integrity of
network-generated data and the continuous operation of the network. Both
SSA and the vendor conduct ongoing tests of the accuracy and

completeness of the network-generated data on which so much of SSA's
800-number related performance measurement, management decisions, and
staffing levels depend. The vendor told us that redundancy was built into
the network to ensure that the failure of any one component only affected
existing calls. For example, if one component fails, the network
automatically employs a backup execution path to bypass the problem
location and reroutes calls to one of the remaining call centers.
According to the vendor, the redundancy built into the 800-number network
and the geographical dispersion of its redundant functions would make a
complete system outage almost unimaginable. Vendor staff told us that the
local outages that occur on occasion are mainly caused by loss of network
facilities, extended local power failures, or hardware issues. SSA and the
vendor maintain back up databases critical to network operations.

SSA takes several additional steps to help ensure that callers can access
800-number services. SSA network operations staff frequently calls the
800-number network to test the integrity of the main menu scripts and the
routing of calls to both automated and agent services. They evaluate calls
for proper routing through the option choices; proper functionality of the
automated scripts; proper functionality of routing to agents; proper
routing to agents and agent queues; and the quality and clarity of the
connection. Call centers also have systems administrators who monitor the
performance of the equipment used on the premises and notify headquarters
when any anomalies appear. System administrators are responsible for
keeping the phones and headsets working, troubleshooting problems with
desktop applications, monitoring computers, printers, or management
information data. If the administrators notice any problems, they are
responsible for notifying headquarters so that the vendor can dispatch a
technician to initiate repair.

    SSA Manages Call Traffic and Forecasts Future Call Demand and Staffing Needs

SSA takes advantage of the wealth of management information at its
disposal to monitor ongoing network operations and plan for the future.
SSA forecasts call volumes and schedules the appropriate number of agents
in accordance with anticipated demand based on historical data. These
forecasts allow SSA to group days into specific levels depending on the
anticipated volume of calls. For example, the busiest days-"Level 1"
days-require the greatest number of SPIKES to be activated to answer
phones. SSA sets and tracks SPIKE commitments to help ensure that enough
SPIKES will be available networkwide to answer the volume of incoming
calls. Depending on network conditions, managers may make adjustments to
the number of available agents and the routing of calls to align available
800-number resources with caller demand. SSA adjusted its

call volume forecast downward 5 times each in fiscal years 2003 and 2004,
allowing SPIKES scheduled to answer 800-number calls to return to their
other assigned duties.

SSA uses real-time data to monitor call traffic, caller activity, and
system performance. SSA can use these data to track overall incoming calls
and information on automation or determine whether calls were routed to an
SSA call center or to a busy message. SSA monitors such 800-number network
statistics as calls made to the network, calls offered to agents, agent
staff levels, average speed of answer, and agent busy rate. Staff also
monitors cable and national news for events, such as inclement weather,
news stories on Social Security, or homeland security events to determine
what impact they may have on projected 800-number call volumes.
Furthermore, SSA monitors caller usage of the automated menus and
reshuffles automated options to keep the most popular options first.

SSA performs limited checks of the network generated data. Upon receiving
the data electronically, SSA runs the data through a multistep automated
procedure that backs up the data and converts it to a readable format. As
part of this process, SSA checks each record to ensure that all area codes
are good, all phone numbers are properly formatted, and all listed phone
numbers originate in the 800-number network. The vendor also generates
separate reports on automated services and agents. SSA reviews the reports
and compares the results with historical trends. Although SSA has no
additional means of verifying the reliability of the vendor-provided data
or the results that appear in report field outputs, both SSA and the
vendor maintain that these data are accurate, and the vendor states that
SSA has the source data it needs to assess network performance.

    SSA Expanded Its Automated and Agent-Assisted Services, but 17 Percent of
    Calls Seeking Agent Assistance Do Not Get Through

Since the inception of the nationwide 800 number and the later
introduction of limited 24-hour automated services, SSA has continually
improved the quality and quantity of services available to callers. In
1996, SSA introduced voice-recognition applications and added an option
allowing callers to replace their Medicare card by phone. In 1998, SSA
implemented five new automated service options to handle inquiries
surrounding the increased number of Social Security statement mailings. By
2002, SSA had made the full range of automated services available in the
Spanish language.

Callers may access the automated services at any time in English or
Spanish to obtain services, information, or forms. Table 2 lists the
services available through the 800-number automated menus.

 Table 2: Services Available through the 800-Number Automated Menus, As of June
                                      2005

Main service categories Subcategories

Field office locator database  o  N/A

Password/password request code/account status  o  N/A

Social Security card application form  o  N/A

Benefit verification  o  N/A

Social Security statement options  o  Form to request Social Security
statement

o  Form to correct name or date of birth

o  	Form to request Social Security statement with different estimate of
future earnings

o  Message-why Social Security statement was sent

o  Message-address on Social Security statement

o  Message-earnings on Social Security statement

Medicare options  o  Replacement Medicare card

o  Form to request help with the Medicare Prescription Drug Plan costs

o  Medicare Prescription Drug Plan message

Pamphlets  o  Understanding the benefits

o  Retirement benefits

o  Disability benefits

o  Survivors benefits

o  How work affects your benefits

o  Benefits for children with disabilities

o  What every woman should know

Informational messages  o  Payment delivery dates

o  Direct deposit

o  Best times to call 800 number

o  SSA Internet address and services

o  SSI message

o  1099 benefits statement (seasonal)

o  Cost-of-living adjustment (seasonal)

Source: SSA data.

SSA has adopted the telephone industry "best practice" of taking care of
all of the caller's business during the initial contact. Agents have been
trained to answer a wide range of inquiries and have the capacity to
transfer calls they cannot handle to others who handle these calls. For
example, in 1998, SSA began allowing callers to file claims for retirement

and survivors' benefits immediately through the 800 number, eliminating
the need for the caller to leave a message and wait for another SSA agent
to return the call. In 1999, SSA gave agents access to a computer-based
application to assist them in handling telephone calls more efficiently.
In 2002, SSA provided callers the option of having their call routed to a
designated group of bilingual agents. SSA also extended the hours of agent
availability nationwide. Agents are now available weekdays from 7 a.m. to
7 p.m. in each time zone. In addition, SSA provides unadvertised agent
service for extended hours on weekday nights and weekends. SSA also
provides agent service for the hearing impaired through a separate
tollfree number.

In following SSA's instructions to handle all of the caller's business
needs, agents may be performing tasks that limit their availability to
answer calls. During site visits, we observed agents who filled out forms
by hand, retrieved printouts, placed the mailings in an envelope,
addressed the envelopes by hand, and put the envelope in the mail slot,
while the caller remained on hold. While these steps may help give callers
the assurance that their business is being completed, such manual tasks
are timeconsuming and potentially limit the number of calls that agents
can handle.

Although the number of calls placed to the 800 number has increased
slightly since fiscal year 2002 and SSA has expanded services available
through automation, the agency continues to have difficulty keeping pace
with caller demand for live agent assistance. Figure 3 shows the calls
made to the 800 number since fiscal year 2002 when SSA's most recent
telephone network upgrade was fully implemented. The proportion of calls
to the 800 number indicating a preference for agent assistance has been
relatively consistent, whereas SSA had hoped that the introduction of
automated services would divert calls away from agents to the less costly,
self-service automated system. Such a reduction would be consistent with
the call center industry trend toward self-service with minimal agent
intervention. However, agents continue to answer the majority of calls,
including some calls that, according to agents, could easily be handled
through automation. The percentage of calls seeking agent assistance but
not getting through declined from 22.7 percent in fiscal year 2002 to 15.2
percent in 2003, but rose 2 percent in fiscal year 2004. Specifically, as
figure 4 shows, 8.7 million (or 17.2 percent) of the 51 million calls
seeking agent assistance in fiscal year 2004 did not get through. About
half of these calls encountered a busy message and the other half
abandoned the call while waiting in queue. Managers of private call
centers do not place a lot of importance on call abandonment rates for
several reasons, including

their belief that callers terminate calls to visit the Web site. Some
callers that request SSA agent assistance may be able to satisfy their
needs through the automated menu or Web site. However, if callers'
business require agent assistance, they will not be able satisfy their
needs if they unable to get reach an agent.

Figure 3: Calls Placed to the 800 Number, Fiscal Years 2002 through 2004
(in Millions)

Millions

90

80

70

60

50

40

30

20

10

0 2002 2003 2004 Fiscal year

Calls placed to automation Calls placed to agents Source: SSA data.

Figure 4: 800-Number Calls Seeking Agent Assistance That Did Not Get
Through, Fiscal Years 2002 through 2004 (in Millions)

Millions

8

7 6.5

6

5

4

3

2

1

0 2002 2003 2004 Fiscal year

Calls encountering busy messages

Calls abandoned in agent queue

Source: SSA data.

SSA offers a variety of possible reasons why callers abandon their calls
after being placed in queue for an agent, one being that customers simply
do not want to continue waiting any longer before having an opportunity to
speak to an agent. SSA has several initiatives underway to reduce the
number of abandoned calls in queue, including a call-back service, which
will provide callers kept in queue beyond a certain threshold with an
opportunity to enter their telephone number and select a contact time so
that an agent can call them back. While providing convenience to callers
and potentially using any agent "down" time more efficiently, a call back
option also has the potential to increase agent workload.

Since 2002, SSA's 800-number automated menus have received progressively
higher call volumes but handled fewer calls to completion. In addition, as
shown in figure 5, the number of calls being abandoned without completing
a transaction in the automated menus has steadily increased, culminating
in fiscal year 2004, when nearly half of calls to automation were
abandoned.

Figure 5: 800-Number Calls Handled and Abandoned in Automation, Fiscal
Years 2002 through 2004 (in Millions)

Millions

30

25

20

15

10

5

0 2002 2003 2004 Fiscal year

Calls handled in automation

Calls abandoned in automation

Source: SSA data.

Although SSA offers a number of possible reasons, it is unable to say with
any degree of certainty why calls continue to be abandoned. In the past,
SSA has conducted follow-up caller surveys to ask callers what had
prompted them to abandon the automated services. The primary reason that
callers gave for hanging up after an initial selection of an automated
service was their desire to speak to an agent. According to SSA, many
callers simply desire the security of human contact when leaving personal
information that is required to transact business. SSA has now eliminated
the need for callers to redial; callers may now have their calls
transferred from automated services to agent queue. However, this option
will likely increase agents' call burden. SSA intends to make its
automated menu selections more accessible by introducing a speech-enabled
main menu that would allow callers to simply speak their needs in response
to directed questions. For example, rather than listening to a list of
options, callers will be able to use their voice to narrow down available
options and find the ones relevant to the services they seek. SSA plans to
implement this feature nationwide later in this fiscal year. SSA also
redesigned its Web site in 2003 to improve its accessibility and usability
in the hope of

  SSA Trains and Provides Agents onthe-Job Resources, but Agents Have Not Met
  SSA's Standard for Accuracy of Assistance

relieving the burden on the 800 number. The Web site now attracts over 30
million visitors a year, which SSA says has reduced the demand for direct
service from 800-number and field office agents. SSA's customer
satisfaction surveys from 2002 and 2003 show that the percentage of the
survey respondents who said they would likely use the 800 number the next
time they contacted SSA decreased from 75 to 61 percent. In contrast, the
percentage of respondents who reported they were likely to use the
Internet or email to contact SSA increased by 2 percent and the percentage
of those who said that they would likely call or visit a field office
increased by 10 percent.

SSA has taken steps to help agents provide callers accurate information
and comply with agency requirements, but still has problems with agents
meeting its standards for accurate service. SSA provides agents with
comprehensive training and equips them with on-the-job resources to help
them provide accurate and consistent service. In addition, SSA monitors
agents' calls and compiles agencywide assessments of agent accuracy in
handling calls and identifies agent training needs. SSA's own monitoring
assessments for 1998 through 2003 found that the agency generally met its
standard for agent accuracy in handling issues that had the potential to
affect individuals' benefit payments, but not its standard for "service
accuracy," handling issues that did not have the potential to affect
benefits. SSA's overall performance for "service accuracy" for fiscal year
2003 was 85.1 percent; below SSA's 90 percent target. According to SSA's
assessment, agents' failure to obtain the required identifying pieces of
information from callers to verify their identity before accessing and
disclosing information was the most frequently committed service error. In
fiscal year 2003, this error alone accounted for 28 percent of all service
errors that SSA identified. SSA has taken several actions to help agents
improve their performance, but these actions have not resulted in
sustained improvements in service accuracy.

    SSA Trains Agents and Equips Them with Resources to Properly Handle Calls

SSA provides agents with comprehensive training to enable them to offer
callers a broad range of services and to complete callers' business on
initial contact. The basic training curriculum is comprised of formal
course work to teach agents about the agency's programs, policies, and
procedures,5 including rules for disclosing information to and accepting
reports from callers; how to access, interpret, and enter data into SSA
computer systems and databases; and how to query and interpret SSA
records. As part of their basic training, agents take frequent tests,
conduct mock interviews, observe experienced agents handling calls, and
answer calls. The basic training curriculum for full-time agents at the
call centers we visited ranged from 11 to 16 weeks. In addition, call
center officials told us that they taught a modified 11- to 12-week course
to back-up agents to augment their existing technical skills. In addition,
officials told us that they supplemented the basic training with regional
and call center training offerings, such as new employee orientation,
diversity training, and public service training.

After agents complete basic training, regions and call centers follow
their own established practices to help agents transition to handling
calls on their own. At the sites we visited, agents were mentored or
closely supervised during a transitional period. For example, some call
centers assigned a personal mentor to sit and observe agents handling
calls and to provide prompt assistance, as needed. After spending a number
of weeks with a mentor, agents are evaluated to determine their readiness
to handle calls on their own. As another transitional step, one call
center placed agents in a training unit that had a higher
supervisor-to-agent ratio to allow closer supervision and monitoring of
agents' work. Floor support staff in one training unit said that, in
addition to providing technical assistance, they review the accuracy of
agents' data entries for events, such as direct deposit requests and death
reports. Based on an agent's proficiency, floorsupport staff may review
agents' work to provide daily feedback or review their work less
frequently as agents demonstrate proficiency.

5Guidance for SSA's programs include, the Teleservice Center Operating
Guide (guidance for answering general inquiries and processing events
callers report), the Program Operations Manual System (instructions for
developing actions for input), the Modernized Systems Operations Manual
(instructions for entering transactions into SSA's computer system), the
Social Security Handbook (explains SSA's programs, health insurance, and
related programs), and the Medicare Handbook (explains how Medicare
program services are provided and payments made).

Agents may receive subsequent training in a variety of ways. For example,
training can occur during the 3-hour allotments reserved for monthly staff
meetings. Call center staff and officials told us that these meetings were
used as a forum to provide agents information on emerging issues such as
national and regional initiatives and changes in operating procedures, as
well as feedback on the call center's performance. During the workday,
supervisors may provide agents with important information that agents need
to know, such as generic responses to calls triggered by current media
reports on Social Security solvency. We were told that agents also receive
voluminous intra-agency communications for which they may be allotted 15
minutes at the end of each workday to read. We were also told that
supervisors and floor-support staff use various strategies to ensure that
agents are aware of the most important changes. Call center managers and
supervisors told us, that if needed, more time maybe requested for agents
to be off the telephones to receive additional training, such as hands-on
computer training.

To assist agents in providing callers with accurate and consistent
services, SSA provides agents with the Customer Help and Information
Program (CHIP)-a customized online computer application for providing
services to 800-number callers. CHIP helps agents navigate the
comprehensive set of requirements and guidance for SSA programs and
directs agents in the actions they should take to accurately complete
callers' business on initial contact. For example, if an agent enters an
address change for individuals receiving Supplemental Security Income
(SSI) benefits, CHIP displays screens prompting the agent to ask callers a
series of questions about changes in living arrangements-events that may
lead to an increase or decrease in SSI benefits. As another resource, the
call centers we visited made more experienced staff available to help
agents handle more complex or technical calls. Officials told us that such
floor support was customary at call centers agencywide.

    SSA Monitors Agents to Assess Performance and to Identify Training Needs

SSA monitors agents' handling of 800-number calls for payment accuracy and
service accuracy.6 SSA assesses agent performance for payment accuracy in
cases where agents' responses on such matters as eligibility, filing of
claims, or reportable events could potentially affect an individual's
eligibility or benefits. SSA also assesses agents' performance for service

6SSA's practices for monitoring agents are subject to a memorandum of
understanding with the American Federation of Government Employees.

accuracy to determine whether or not the services they provide correspond
with SSA policies and procedures. When assessing service accuracy, SSA
considers whether agents provided accurate information as well as
performed all other related actions that the agency requires. Some of
these actions are required as a matter of convenience to callers or to
avoid the potential need for follow-up contact.

SSA conducts random, remote monitoring of agents handling calls for
various purposes. OQA is responsible for two types of monitoring. First,
OQA monitors a statistical national sample of calls handled by agents
throughout the year to develop both agencywide and regional assessments of
agent performance. This type of monitoring serves as SSA's means of
assessing agent payment accuracy and service accuracy. OQA officials told
us that such monitoring had the capacity to reveal issues that needed to
be addressed at the agency level, such as pinpointing areas needing policy
clarification. However, the responsibility for agent performance,
including improving performance to meet agency targets, rests with the
various regions and individual call centers. Second, if requested by
regional officials, OQA occasionally monitors a small number of calls
handled by individual call centers and visits the call centers to brief
managers and agents on its findings. Call center staff also randomly
monitor calls handled by their call center for payment accuracy and
service accuracy and to identify training needs for their agents. SSA does
not specify the number of calls that should be monitored for this purpose.
Call center officials told us that the number of calls they monitored do
not provide a statistically valid assessment of their center's
performance.

Designated call center personnel also monitor individual agents to provide
agents individualized feedback on their telephone performance. Monitors
may point out positive aspects of agents' performance as well as suggest
additional training. Agents are given advance notice of when monitoring
will occur and are allowed to choose whether to have monitors sit with
them or to have monitors listen in from a remote location.7 For full-time
agents, SSA guidance recommends monitoring as many as five calls per month
for agents with more than 1 year of experience and unlimited calls for
agents with less than 1 year.8 Officials told us that agents are given

7The purpose of monitoring individual agents is to assess an agent's
performance and provide feedback without creating a punitive environment.

8The national American Federation of Government Employees agreement allows
SSA to monitor all full-time agents individually but requires SSA to
bargain with local unions to similarly monitor backup agents.

timely feedback on assessments of their overall performance. Some
officials also said that when monitors observe agents making an error,
they may interrupt the call to instruct the agent on the correct
procedure.

    Agents Have Not Met SSA's Standard for Accuracy of Assistance

Although SSA takes a number of actions to help agents provide callers
accurate information in accordance with agency policies and procedures,
agents still have problems meeting SSA's standard for service accuracy. As
shown in figure 6, from fiscal year 1998 through fiscal year 2003, SSA
generally met its 95 percent target for payment accuracy-having agents
correctly handle inquiries involving eligibility and benefit payment
issues-but not its 90 percent target for service accuracy-having agents
serve calls related to nonpayment -related issues according to agency
requirements.9

9In fiscal year 2003, SSA discontinued externally reporting performance
for meeting the goals it set for payment accuracy and service accuracy.
However, the agency has indicated that it remains committed to agents
delivering an acceptable level of service. SSA officials told us that the
agency continues to use the performance goals internally.

Figure 6: SSA's Performance in Meeting Payment Accuracy and Service
Accuracy Targets, Fiscal Years 1998 through 2003

Percent

100

95

Payment accuracy target

90

Service accuracy target

85

80

75

0 1998 1999 2000 2001 2002 2003

Fiscal year

Payment accuracy performance

Service accuracy performance

Source: SSA weighted estimates based on sample data.

Note: OQA does not routinely calculate and publish confidence intervals at
the agency level. However, OQA calculated the approximate sampling
variability for the fiscal year 2003 results at the 95 percent confidence
level to be +- .8 percent for payment accuracy for the universe of calls
with a payment-affecting issue and +-1.1 percent for service accuracy. OQA
said that because the call population, the number of calls monitored, and
accuracy rates have remained relatively constant for the period of our
review, the confidence intervals would change by only tenths of a
percentage point from one year to the next.

SSA reported that its overall performance for payment accuracy in fiscal
year 2003 was 95.9 percent, and the performance for each of its 10 regions
was similar. However, SSA reported its overall performance for service
accuracy in fiscal year 2003 was 85.1 percent. Based on OQA's assessment,
as few as four regions may have met the 90 percent service accuracy target
in 2003.10 As shown in figure 7, for fiscal years 2001 through 2003,
almost all regions had problems consistently meeting SSA's established
target for service accuracy.

10Given that OQA estimated the sampling error for fiscal year 2003 could
have been as large as +-4.4 percent, only the four regions for which OQA
reported service accuracy rates of at least 85.6 percent may have met the
90 percent service accuracy target.

Figure 7: Regional Performance for Meeting Service Accuracy Target, Fiscal Years
                               2001 through 2003

Percent

                            Service accuracy target

Boston

kr

                              Philadelphia Atlanta

goChica

                               Dallas Kansas City

verDen

                            San Francisco SeattleYo

wNeRegion

2001

2002

2003

Source: SSA weighted estimates based on sample data.

Note: OQA does not routinely calculate and publish confidence intervals at
the regional level. OQA estimates the sampling variability for FY 2003 at
the 95 percent confidence level, would range from +-2.9 percent to +-4.4
percent for service accuracy. OQA said it would expect similar regional
sampling variability for prior years.

OQA identified 63 types of required actions that agents failed to take in
fiscal year 2003 that led SSA to miss its service accuracy target. Agents'
failure to take these required actions resulted in service errors. As
shown in table 3, the most frequent error stemmed from agents' inadequate
protection of individuals' personal information. SSA protects individuals'
privacy by limiting disclosure of the personal information in its records
to individuals for whom the agency maintains the records and to others
authorized. Agents committed an error each time they failed to collect the
requisite six identifying pieces of information to verify a caller's
identity

before accessing or disclosing information from SSA records (i.e.,
improper handling of access and disclosure).11

     Table 3: Top Five Types of Service Errors Observed in Fiscal Year 2003

                                                    Percentage of all service

Type of service errora errors Example of service error

Caller inquiry and agent action Action agent failed to take

Improper handling of 28 A caller representing an attorney's office wanted
to

access/disclosure	know when a client who was in the attorney's office
should expect payment of disability benefits. The client had received a
letter indicating SSA had made a favorable decision.

The agent obtained the client's Social Security number from the caller,
verified that a favorable decision was made, and told the caller that it
usually takes 3 months from the date of the letter to receive payment.

Failure to obtain or 10 A caller wanted to know the status of her child's
properly interpret a query disability claim.

The agent misread SSA's records, advised the caller that the claim was
denied, and offered to mail appeal forms.

Improper handling of 8 A widow reported her husband's death to stop his

b

potential claims	retirement checks. The agent accepted the death report
and told the widow she should return the husband's next payment to her
local SSA office.

Improper handling of 8 Caller wanted to obtain replacement of a lost
Social referral to another SSA Security card. The caller acknowledged
having two of component the identity documents that the agent said she
would Agent should have asked to speak to the client to verify client's
identity and ask the client's permission to conduct business with the
attorney representative.

Agent should have advised the caller that a final decision had not been
made and to allow 60 days for SSA's decision.

Agent should have screened the widow to determine her possible eligibility
for benefits on her husband's account.

Agent should have offered to mail the caller a Social Security card
application form that would have allowed caller to obtain a replacement
Social Security card without having to visit an SSA office.

need to get a replacement.

Agent told the caller to contact the SSA office ahead of her visit to make
sure the two documents were sufficient and was given the SSA office
address and telephone number.

11SSA does not consider the ability to provide all the required
identifiers to be proof of callers' identity. Similarly, the agency does
not consider the failure to provide required identifiers as conclusive
evidence that callers are not who they allege.

                                                    Percentage of all service

Type of service error a errors Example of service error                    
                                6 Caller was payee for two children receiving 
Improper handling of                                             benefits, 
acceptable reporterc           one of whom recently turned 18 and was away 
                                  at                                          
                                  school. Caller wanted to report address     
                                  change for                                  
                                  both children.                              

Agent properly identified the caller, processed a change of address for
the younger child, and told the caller the older child would have to
report the address change herself.

Agent should have accepted address change for both children because caller
was an acceptable reporter.

Source: OQA "800 Number Evaluation Reports" for fiscal years 2001 through
2003.

Note: The examples provided are for illustrative purposes and the actual
incident depicted may not have occurred in fiscal year 2003.

aOQA identified 58 additional causes of service errors, 9 of which
occurred at frequencies between 1 and 5 percent, and the remaining 49
occurred at a frequency of less than 1 percent.

bInquiries about claims where there was no reasonable probability to
affect payment or eligibility for benefits.

cThis error occurs when agents either accept a nonpayment-affecting report
without verifying the identity of the reporter to determine if the report
can be accepted, or refuses to accept a nonpaymentaffecting report from a
proper reporter.

Managers at the sites we visited have taken actions to reduce the number
of service errors, particularly access and disclosure errors. For example,
some call centers provided CHIP refresher training, designed desk aids
reminding agents of the steps for proper disclosure, placed "hot pink"
sheets detailing the service errors on the desks of agents who commit
them, or established a "CHIP doctor" to provide agents with technical
assistance to help navigate the CHIP computer application. However, the
effectiveness of these actions to improve service accuracy for agents
within the respective call center is unknown because the monitoring that
occurs at the call center level does not provide a statistically valid
measurement to make such an assessment.

OQA has reported that the lower service accuracy rate primarily stemmed
from agents' failure to follow SSA's requirements when asking callers to
verify their identities. Assuming that such "access and disclosure"
failures could be cleared up through the use of the CHIP application, SSA
mandated its use in November of 2001. The service accuracy rate
subsequently improved for fiscal year 2002, but dropped the next fiscal
year because, according to OQA, agents did not make optimal use of CHIP.
OQA recomputed SSA's service accuracy rate without the access and
disclosure error for comparison purposes and reported that it would have
increased from 85.1 to 89.2 percent in fiscal year 2003. SSA has not

  SSA Conducts Training, Monitoring, and Customer Surveys but Lacks a Uniform
  System for Assessing Agent Courtesy

determined why agents fail to follow agency procedures when handling some
calls, resulting in service errors.

SSA uses training, call monitoring, and surveys to ensure that agents
deliver courteous service, but does not routinely document or analyze all
incidents of discourtesy or caller complaints. As part of its
comprehensive, multiweek training curriculum, SSA teaches agents the
interviewing and interpersonal skills they need to provide courteous
service. It also determines through monitoring whether agents are being
courteous. Based on its monitoring results from 2001 through 2003, SSA
reported that it found agents to be courteous to callers over 99 percent
of the time. SSA also measures caller satisfaction with agent courtesy as
part of its annual 800-number customer satisfaction survey. In 2004, 91
percent of respondents rated agent courtesy as good, very good, or
excellent; 5 percent rated agent courtesy as fair, and 4 percent rated it
poor or very poor. While SSA uses training, monitoring and customer
surveys to ensure courtesy, it does not have a uniform system for
analyzing incidents and complaints of discourtesy. Call center staff told
us that they typically apologize to callers and offer to provide the
desired assistance whenever callers lodge complaints by phone. Moreover,
they may not record the complaint or attempt to capture and assess
information on the nature of complaints. Customer service studies
highlight the importance of paying attention to complaints and the
benefits of having a good management complaint system.

    SSA Relies on Training, Monitoring, and Customer Surveys to Ensure Agent
    Courtesy

As part of its comprehensive, multiweek training curriculum, SSA teaches
interviewing and interpersonal skills to help agents serve callers in a
professional and courteous manner. The training includes instruction on
how to establish rapport with callers, how to obtain information necessary
to accurately serve callers' needs, and how to end calls on a positive
note. As a courtesy measure, agents are instructed to allow callers to end
the call. Agents also receive training on how to respond to angry, loud,
or abusive callers, including how to calm such callers, and how to
continue serving them or to transfer those calls to supervisors.

SSA also uses its call monitoring process to oversee courtesy levels and
has procedures for immediate intervention to remedy any observed

problem.12 OQA procedures call for monitors to immediately inform
management of a discourteous incident, prepare a written report for the
agent's call center manager, and retain a copy of the report in the event
that a disciplinary action is taken against the agent. Call center
managers who become aware of discourtesy allegations or observe agent
discourtesy are required to follow similar procedures. They are required
to discuss any incident with the agent and consider a progressive range of
disciplinary actions from issuing a reprimand to terminating an agent's
employment.

OQA officials told us that formal monitoring is time-consuming work. As a
result, OQA said that over the years, it reduced the sample size of the
monitored calls due to resource constraints. Regional and call center
management expressed varied opinions as to whether the reduction in the
number of monitored calls was an obstacle to identifying agent
discourtesy. One call center manager told us that discourtesy was more
likely to be observed by managers and supervisors patrolling work areas
than through formal monitoring. On the other hand, one regional official
noted that additional unannounced monitoring would be a more effective way
of catching agent rudeness.

Some of the managers and officials with whom we met, however, told us that
they believed courtesy levels were very high and not a problem. According
to the agency's call monitoring records, SSA agents have performed at
consistently high rates with regard to courtesy. For fiscal year 2003, OQA
determined that based on 4,384 calls, agents had been courteous to callers
99.9 percent of the time. 13 It reached similar conclusions from its 2001
and 2002 monitoring.

SSA also relies on its annual survey of callers to assess and ensure agent
courtesy. Callers who have used agent services have been asked, among
other questions, to rate agent courtesy on a 6-point scale. The 2004
survey showed that 91 percent of the callers rated agent courtesy as being
good, very good, or excellent; 5 percent rated it as being fair; and 4
percent rated

12OQA defines courtesy as a reflection of whether agents treated callers
in accordance with generally accepted standards. Other SSA guidance
provides examples of unacceptable agent conduct, including hanging up on
callers without cause, using profanity, yelling at callers, exhibiting
rudeness or impatience with callers, using inappropriate tone, and putting
callers on hold unnecessarily.

13OQA observed six calls of discourteous service that they said
represented a population of nearly 60,000 calls.

it as being poor or very poor.14 These rates were about the same as those
reported for the 2001 though 2003 surveys.

Other call centers may use telephone or online surveys to obtain feedback
from customers, although the actual administration of the surveys may
vary. For example, one organization conducts telephone surveys using voice
capture software to record customer responses. At the beginning of a call,
the survey system randomly selects participants and asks them to
participate in a 2-3 minute survey after they complete their call. Another
organization conducts online surveys, sending a survey to selected
customers via email. Nonetheless, these organizations seek to obtain
customers' views on their organization's performance.

    SSA Lacks a Uniform System for Assessing Agent Courtesy

SSA monitors calls and receive feedback from customers, but it does not
systematically gather and assess this information to identify courtesy
problems, such as particular problem locations or persistent patterns or
trends. SSA agents handled an average of 40.9 million calls each year from
2001 through 2003. Even if agents were courteous 99.9 percent of the time
as OQA reported, for fiscal year 2003 that would still leave nearly 60,000
calls in which the agents may have been discourteous. However, because SSA
does not routinely analyze the details of agent discourtesy observed
through monitoring, it has no way of determining the circumstances or
lessons learned from monitored calls.

Studies conducted on customer service have shown that building
relationships with customers and a having first-rate management complaint
system are critical to maintaining good customer relations. One study in
particular noted that paying attention to customer complaints, regardless
how minor they may be and addressing them quickly and completely helps
satisfy customers and build trusting relationships. 15

14A study contracted by the Office of the Inspector General reported that
SSA's customer satisfaction survey provided a reliable statistical
representation of caller's views for the period measured. However, because
SSA limits the survey to being conducted during a designated period-most
recently a 4-week period in March-the survey may be an inadequate
representation of customers' views year round. Further, the sampling
variability for responses to this question would be +-2 percent or less at
the 95 percent confidence level. SSA's response rate during the period
2001 through 2004, ranged from 53 to 71 percent. Although response rates
within these ranges are not unexpected for this kind of telephone survey,
it should be noted that as the response rate decreases, the certainty that
survey results represent the universe of 800-number callers decreases.

15Richard D. Young, Customer Driven Focus and Excellence in the Public
Sector (Columbia, S.C.: Institute for Public Service and Policy Research,
University of South Carolina, 2002).

Similarly, people who contact their government agencies want to be heard
and expect courteous and respectful treatment. It is therefore important
for government employees to distinguish what their customers want and to
take actions to ensure that their customers are satisfied. The study also
noted that no matter how good the service or product is, occasions will
invariably arise that result in customer complaints. However, it is
important that when criticisms are voiced, that they are systematically
and promptly addressed. A good management complaint system can provide
data and information on complaints that can be compiled and analyzed to
give insight into where problems are recurring and what needs to be done
to fix them or prevent them from happening in the future. A good complaint
system also facilitates the filing of complaints using simple, yet
comprehensive complaint forms.

SSA's 800-number customer satisfaction surveys are one means of gathering
feedback from callers on agent courtesy. However, the survey does not ask
why some respondents rate agent courtesy as poor. In addition, the agency
does not routinely collect or analyze all caller complaints placed through
the 800 number. Our visits to call centers found variation in how they
handled such calls. When customers call the 800 number to report
agent-related complaints, SSA guidance requires agents to refer calls to
supervisors or floor-support staff. However, SSA does not provide guidance
for how those receiving referrals should handle them. We were frequently
told that call center staff receiving these calls typically apologize for
the other agents' rudeness and offer to provide service to the caller. SSA
provides call center staff a form to document 800-number service
complaints, including agent lack of courtesy. However, SSA has not
provided them agencywide guidance on documenting complaints or the type of
information they should record to allow SSA to identify service issues or
trends. We were given a variety of reasons why call center staff may not
document agent-related complaints. One call center official told us that
his site allowed agents to exercise judgment in deciding which complaints
they documented. Some agents, supervisors, and technical staff told us
they were unaware of procedures for handling such complaints while others
believed callers needed to provide sufficient information such as the
offending agent's name or call center location to lodge a formal
complaint. It was our observation that 800-number agents may not provide
their full name or mention their call center location when answering a
call.

SSA responds differently to customer-reported complaints sent to agency
offices than to complaints registered on its Web site. Specifically,
regional and call center officials said that when warranted they would
attempt to

Conclusions

identify the agent, investigate the merits of the complaints reported to
their offices, and initiate disciplinary actions. Headquarters staff who
receive complaints through the agency's Web site told us that they
routinely send customers a letter of apology, but have no one designated
within SSA to whom to forward and or resolve complaints. Although the Web
site has an Intranet-based form ostensibly designed to capture complaint
information, it does not ask for specific information such as the nature
of the alleged act of discourtesy and the date and time it occurred. By
not systematically collecting and analyzing information on alleged agent
discourtesy, SSA is unable to identify service issues that may warrant
corrective actions.

SSA's toll-free phone service is an important resource for the thousands
of people who call the number daily, and the steps the agency has taken in
recent years demonstrate a commitment to quality service. The agency's
upgraded telephone system along with its expanded menu options and
agent-assisted services has improved access in many respects by giving
customers more services at their convenience. In addition, the agency has
taken steps to help ensure that callers receive more accurate and
courteous service.

Even with good service, however, there is room for improvement.
Improvements to the 800-number network have not necessarily ensured that
callers receive the help they seek, given the number of calls not reaching
an agent. This may not be a mounting problem if callers follow the general
trend in the call center industry toward automation and selfservice as
they grow more comfortable with these options. However, the impending
increase in the size of the retiree and disability populations, and
anticipated changes to the Social Security system suggest that SSA may
continue to experience a substantial proportion of callers who request
agent assistance. Measures to improve customer access to agents may
therefore be needed. In addition, SSA's many benefit programs will
continue to generate some complex questions that require agent assistance.
Currently, the prevalence of service accuracy errors diminishes the
quality of service that callers receive when they do reach an agent
through the 800 number. Finally, although SSA's estimates show instances
of agent discourtesy to be rare among all calls, such instances could
nonetheless affect tens of thousands of callers. Because SSA does not
routinely capture information on all customer complaints about
discourtesy, however, it loses the ability to assess the severity of the
problem and misses opportunities to better understand caller needs, solve
unanticipated problems, and retain the good will of the public.

  Recommendations for Executive Action

Agency Comments and Our Evaluation

To improve the quality of the 800-number telephone service, we recommend
that the Commissioner of the Social Security Administration take the
following three steps:

o  	Identify cost-effective ways that will help ensure that more calls
seeking agent assistance get through to agents, such as streamlining the
call-handling process, automating some mailings that agents now do by
hand, or increasing number of agents available to take calls.

o  	Conduct a comprehensive analysis of the source of service errors. For
example, the agency might consider holding agent focus groups to gain
insight into why agents tend to fail to comply with certain requirements.
The agency could get agents' views on the effectiveness of CHIP in helping
them meet agency requirements.

o  	Establish procedures for documenting and assessing customerreported
complaints. In doing so, the agency should determine the types of
information it needs to assess customers' concerns and to provide the
agency a means to identify and address service issues.

We obtained written comments on a draft of this report from the
Commissioner of SSA. In its comments, SSA said it was pleased that our
report reflected the agency's commitment to providing high-quality
800-number telephone service that meets the needs and expectations of its
customers. SSA agreed with our recommendation to identify cost-effective
ways to increase agent availability to handle 800-number calls and
described several planned initiatives to improve agent productivity and to
expand automated services. SSA also agreed with our recommendation to
conduct a comprehensive analysis of the source of agent service errors.
Accordingly, SSA said it would convene a workgroup to obtain feedback on
the source of agent service errors and make recommendations as appropriate
to improve the agency service accuracy level.

SSA disagreed with our recommendation to establish procedures for
documenting and assessing customer complaints. SSA said that its findings
that agent courtesy levels are consistently high demonstrate that its
present approach to ensuring agent courtesy-which combines training,
monitoring, and customer surveys-is working. Moreover, SSA said that based
on its experience with prior initiatives, a nationwide reporting system
would require heavy resource expenditures and be cost prohibitive given
current budget constraints. Furthermore, SSA stated that any use of agent
resources to document complaints would be counterproductive to improving
caller access to agent services.

While we agree that agent courtesy levels are high and state this in the
report, given the sheer volume of 800-number calls SSA receives, even
relatively small percentages of callers encountering agent discourtesy
could result in tens of thousands of callers not getting the service they
deserve. Thus, we believe that SSA can benefit from having uniform
procedures for documenting and assessing customer complaints and have
added additional information for further clarification. Experts believe
that paying attention to customer complaints, however minor, and working
to quickly resolve them is important to building relationships with
customers. In addition, having information on complaints helps identify
recurring problems and potential fixes as well as help prevent future
occurrences. Under SSA's current practices, because the decision to
document a complaint lies with the individual agent handling the call,
customers contacting the 800 number have no assurance that SSA will review
the merits of their complaints. Routinely documenting and assessing
customer-initiated feedback could help the agency identify areas of
concern to callers and reinforce the agency's commitment to providing
quality "citizen centered" service.

While we understand SSA's concerns about resource constraints, we maintain
that SSA can implement a system to document complaints using existing
mechanisms, such as its 800-number feedback form and Internet form for
complaints reported to its 800 number and Web site, respectively. As we
state in the report, SSA already devotes time and staff to the
documentation and handling of customer-reported complaints; however, such
efforts are not done routinely. SSA states that its agents provide more
efficient service when they keep the caller on the phone until the
caller's business and all agent actions are completed. We believe
routinely documenting callers' concerns would take no more time than
completing callers' other business. Further, the information could be
collected uniformly in an electronic format that would facilitate analysis
that could be used to improve service. As others have pointed out, a good
system for managing complaints should be comprehensive, yet simple.
Finally, we believe that understanding and responding to customer
complaints are integral to the delivery of quality customer service.

SSA's comments are reproduced in appendix II. SSA also provided technical
comments, which we have incorporated in the report as appropriate.

As agreed with your offices, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days
after its

issue date. At that time, we will send copies of this report to the
interested
congressional committees and the Commissioner of SSA and will make
copies available to others upon request. In addition, the report will be
available at no charge on GAO's Web site at http://www.gao.gov.

If you or your staff have any questions about this report, please contact
me
at (201) 512-7215 or [email protected]. Contact points for our Offices of
Congressional Relations and Public Affairs may be found on the last page
of this report. Key contributors to this report are listed in appendix
III.

Barbara D. Bovbjerg, Director
Education, Workforce and Income Security Issues

Appendix I: Objectives, Scope, and Methodology

The objectives of this study were to evaluate SSA's actions for ensuring
callers have ready access to 800-number services and receive accurate and
courteous service from agents. To do this, we reviewed published works
that included the National Performance Review benchmarking reports to
identify industry benchmarks in areas key to our work and issues
surrounding call center services.1 We also reviewed GAO and the Office of
the Inspector General (OIG) reports and SSA annual performance plans to
identify what is currently known about SSA telephone service operations.
To evaluate the quality of the 800-number service, we compared telephone
system performance data compiled by a contractor for SSA and SSA's Office
of Quality Assurance and Performance Assessment (OQA) assessments of agent
accuracy and courtesy to SSA's established standards and, where
applicable, to industry benchmarks. We used performance data from OQA's
monitoring of agents for fiscal years 1998 through 2003 and from OQA's
800-number customer satisfaction surveys conducted fiscal years 2001
through 2003. We reviewed OQA's management reports for these activities.

To develop information on the actions SSA takes at the headquarters level
to ensure quality 800-number telephone service, we reviewed documents
related to (1) SSA's forecasts of call volumes and projected staffing
levels for auxiliary agents; (2) services offered using the automated
menu; (3) vendor-contracted services for the 800-number telephone systems
hardware, software, and performance data; and (4) requirements for
training agents, monitoring agent performance, and agent courtesy to
callers. We interviewed officials in the Office of Telephone Services to
obtain an understanding of the general operation of the 800-number
telephone system including the routing of calls; the compilation of
performance data; and SSA's actions to monitor the performance of the
800-number system and of the vendor. We also interviewed OQA officials to
obtain more detailed information on procedures for monitoring agents and
surveying 800-number callers. In addition, we reviewed some complaints
reported by the public over the agency's Web site and interviewed
officials in the SSA's Center for Program Support to discuss practices for
handling complaints.

1National Performance Review, "Serving the American Public: Best Practices
in Telephone Service," National Performance Review Benchmarking Study
Report (Washington, D.C.: 1995).

Appendix I: Objectives, Scope, and Methodology

We visited six call centers to observe the 800-number service operations
at the regional and call center levels. At the locations visited, we
observed officials monitoring their centers' call traffic and agent
availability in real time, officials monitoring agents handling live
calls, and agents handling live calls from customers. We reviewed
documentation call center officials provided on agent training, monitoring
of agents, agent-related complaints received, and disciplinary actions
taken against agents. We interviewed regional and call center officials
having line-management, supervisory, floor support, monitoring, and
call-handling responsibilities to obtain information on call center
operations and their experiences in providing telephone services and
serving the public. The call centers we selected varied in the frequency
and volume of calls they handled-three handled calls routinely and three
on a back-up basis-and are not representative of call centers SSA-wide.

To assess caller access and the reliability of the 800 number, we
interviewed SSA officials and contacted selected vendor staff to obtain
documents and data on the 800-number management and operations. SSA uses
the management data and information supplied by the vendor to track all
calls and transactions on the network, including data on overall incoming
calls and information on automation and to determine whether calls were
routed to an SSA call center or to a busy message. The vendor's reporting
system has internal alarms running on each server and an application that
periodically checks each server's vital functions, capacity, and
environmental operating conditions against a predetermined set of normal
operational conditions. Upon receipt, SSA runs the vendorsupplied data
through a multistep automated procedure that backs up the data, creates
data storage files, extracts data to be stored in other datasets, and
recreates the data in a readable format. As part of this process, SSA
checks each record to ensure that all area codes are good, all phone
numbers are properly formatted, and all answering telephones originate in
the 800-number network. The vendor also generates separate reports on
automated services and agents. SSA reviews the reports and compares the
results with historical trends. Although SSA has no additional means of
verifying the reliability of the vendor-provided data or the results that
appear in report field outputs, both SSA and the vendor maintain that
these data are accurate and the vendor states that SSA has the source data
it needs to assess network performance. We reviewed SSA performance data
related to access and determined the data to be sufficiently reliable for
the purposes of this report.

To assess the reliability of OQA's monitoring assessments of agents'
performance, we examined data reliability issues identified in an OIG

Appendix I: Objectives, Scope, and Methodology

report and interviewed OQA officials knowledgeable about the monitoring
process and resulting data. In addition, we reviewed documentation and
training materials, including monitoring instructions, evaluation data
entry forms and desk aids, corrective action and evaluation feedback
forms, and information regarding the statistical sampling of calls. In
evaluating OQA's sampling and weighting methodology, we determined that
OQA's methodology for monitoring agents' payment and service accuracy
appears to adequately represent the population of telephone calls.
Approximate confidence intervals were produced by OQA using standard
formulas for proportions based on a simple random sample. As OIG
previously reported, 2 we also found that decisions regarding payment
accuracy and service accuracy continue to be unverifiable because SSA does
not maintain documentation of all monitored calls. We determined that the
data were sufficiently reliable for our purposes, given these limitations.

To assess the reliability of the survey of 800-number callers, we
interviewed OQA officials about the survey and resulting data and reviewed
documentation on the survey methodology, sampling, response rates, and
sampling variability. We also reviewed a report contracted by the OIG
regarding this measurement of customer satisfaction.3 This report
concluded that the 800-number caller survey produced a reliable
measurement of callers' views of agent courtesy for the period measured,
but that because the survey was administered only twice a year, it was
unlikely that the survey results matched the true customer satisfaction
across the entire year. Because the survey was recently limited to being
conducted during a single 4-week period in March, we found that the survey
results continue to be unrepresentative of callers' responses throughout
the year. We believe that seasonal events could affect customer
satisfaction in different ways throughout the year. The survey response
rate during the period 2001 through 2004, ranged from 53 percent to 71
percent. Although response rates within these ranges are not unexpected
for this kind of telephone survey, it should be noted that as the response
rate decreases, the certainty that the survey results represent the
universe decreases. We determined that the survey data are

2Social Security Administration Office of the Inspector General,
Management Advisory Report: Performance Measure Survey of the Percent of
800-Number Calls Handled Accurately, OIG-A-08-01-11024 (Washington, D.C.,
August 2001).

3Social Security Administration Office of the Inspector General,
Performance Indicator Audit: Customer Satisfaction, OIG A-02-02-11082
(Washington, D.C.: February 2003).

Appendix I: Objectives, Scope, and Methodology

sufficiently reliable for providing a general indication of customer
satisfaction, for the specified periods of administration.

We conducted our work at SSA headquarters, Baltimore, Maryland; at
regional offices in Birmingham, Alabama; Kansas City, Missouri; and
Richmond, California; and at two call centers in each region. We conducted
our work from September 2004 through July 2005 in accordance with
generally accepted government auditing standards.

Appendix II: Comments from the Social Security Administration

Appendix II: Comments from the Social Security Administration

Appendix II: Comments from the Social Security Administration

Appendix II: Comments from the Social Security Administration

Appendix II: Comments from the Social Security Administration

Appendix III: GAO Contact and Staff Acknowledgments

GAO Contact

  Staff Acknowledgments

(130414)

Barbara D. Bovbjerg, Director (202) 512-7215

The following individuals made important contributions to this report:
Shelia Drake, Assistant Director, Jacquelyn Stewart, Analyst-in-Charge,
Susan Bernstein, Michelle Fejfar, Jonathan McMurray, and Roger Thomas.

  GAO's Mission

Obtaining Copies of GAO Reports and Testimony

The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's Web site (www.gao.gov). Each weekday, GAO posts newly
released reports, testimony, and correspondence on its Web site. To have
GAO e-mail you a list of newly posted products every afternoon, go to
www.gao.gov and select "Subscribe to Updates."

Order by Mail or Phone 	The first copy of each printed report is free.
Additional copies are $2 each. A check or money order should be made out
to the Superintendent of Documents. GAO also accepts VISA and Mastercard.
Orders for 100 or more copies mailed to a single address are discounted 25
percent. Orders should be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: 	Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

  To Report Fraud, Contact:

Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected] Programs Automated answering system: (800)
424-5454 or (202) 512-7470

Gloria Jarmon, Managing Director, [email protected] (202)
512-4400Congressional U.S. Government Accountability Office, 441 G Street
NW, Room 7125 Relations Washington, D.C. 20548

Public Affairs 	Paul Anderson, Managing Director, [email protected] (202)
512-4800 U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548

                           PRINTED ON RECYCLED PAPER
*** End of document. ***