Summary

Employment Development Department: Although New Telephone Services Have Enhanced Customer Access to the Department’s Unemployment and Disability Insurance Programs, Customers Encounter Difficulties During Peak Calling Periods July 2001 99031 rotiduA etatS ainrofilaC S T I D U A E T A T S F O U A E R U B The first five copies of each California State Auditor report are free. Additional copies are $3 each, payable by check or money order. You can obtain reports by contacting the Bureau of State Audits at the following address: California State Auditor Bureau of State Audits 555 Capitol Mall, Suite 300 Sacramento, California 95814 (916) 445-0255 or TDD (916) 445-0255 x 216 OR This report may also be available on the World Wide Web http://www.bsa.ca.gov/bsa/ Alternate format reports available upon request. Permission is granted to reproduce reports. C S A ALIFORNIA TATE UDITOR ELAINE M. HOWLE STEVEN M. HENDRICKSON STATE AUDITOR CHIEF DEPUTY STATE AUDITOR July10, 2001 99031 TheGovernor ofCalifornia President pro TemporeoftheSenate SpeakeroftheAssembly StateCapitol Sacramento, California 95814 DearGovernorandLegislativeLeaders: As required byChapter 329, Statutes of 1998, the Bureau of State Audits presents its audit report concerning the Employment Development Department’s (department) introduction of toll-free telephone services to enhance customer access to its unemployment insurance (UI) and disability insurance(DI)programs. This report concludes that the department’s efforts have improved customer service and increased public access to its UI and DI programs. Despite these efforts, during peak calling periods customers may encounter busy signals, hear instructions to call back later, or endure lengthy wait times if they ask to speak with a customer service representative. Also, the department cannot currently measure whether either of the programs meets the goals it established to improve the response time to customers. Specifically, the department cannot distinguish the wait time of customer information calls from claim-filing calls in its UI program. Additionally, the department evaluates the performance of its DI program using management reports that do not routinely include the program’s customer service units, which receive 42percent oftheprogram’scalls. Respectfullysubmitted, ELAINEM.HOWLE StateAuditor BUREAU OF STATE AUDITS 555 Capitol Mall, Suite 300, Sacramento, California 95814 Telephone: (916) 445-0255 Fax: (916) 327-0019 CONTENTS Summary 1 Introduction 7 Chapter 1 Unemployment Insurance Customers Receive Improved Services Over the Telephone but May Endure Long Wait Times During Peak Calling Periods 15 Recommendations 29 Chapter 2 Despite the Department’s Efforts to Improve Access to Program Information, Problems Hamper the Disability Insurance Program’s Customer Service 31 Recommendations 39 Appendix Summary of New Telephone Services’ Effects on the Unemployment Insurance and Disability Insurance Programs 41 Response to the Audit Health and Human Services Agency 43 SUMMARY RESULTS IN BRIEF I n the late 1990s the Employment Development Department (department) introduced toll-free telephone services to enhance customer access to its unemployment insurance Audit Highlights . . . (UI) and disability insurance (DI) programs. Our review indicates that the department’s efforts have improved customer service Our review of the Employment and increased the public’s access to the programs. However, Development Department’s during peak service periods, callers may encounter busy (department) introduction of toll-free telephone services signals, hear instructions to call back later, or endure lengthy for customers of its wait times if the customers ask to speak to a customer service unemployment insurance and representative. disability insurance programs reveals that: With a mid-1990s shift in its methods for helping customers, (cid:1) Its efforts have improved the department has improved the claim-filing process for the customer service and UI program, which the State established in 1938 to compensate increased the public’s access to the programs. individuals for lost wages. The department no longer requires customers of the UI program to request assistance in person; (cid:1) Customers of both instead it offers a spectrum of services over the telephone. The programs are generally toll-free telephone services have also boosted for customers the satisfied with the services. convenience of obtaining assistance. For example, the depart- (cid:1) Despite its efforts, callers ment estimated in 1995 that a person could file a claim over may encounter busy the telephone in as little as 15 minutes, which is a significant signals, hear instructions to call back later, or decrease from the department’s estimate that filing a claim in endure lengthy waits person required an average of 2 hours and 48 minutes of an if they ask to speak individual’s time. Our review disclosed that from April 2000 to a service representative through March 2001, the average call length for filing a claim or during certain periods. for acquiring information from a customer service representative (cid:1) The department cannot was approximately 9 minutes. measure whether the programs have met the The new telephone services also benefit the UI program and its goals established for desired response times to customers in other ways. Callers can obtain some types of their customers. information—including the status of their unemployment checks—seven days a week without ever speaking to UI program staff. Moreover, the department’s improved telephone services allow it to use program resources more efficiently than it did previously. For example, since the department implemented a process that links its six call centers for the UI program, cus- tomer service representatives spend little time idle. Further, with the linking of call centers, customers have access to a larger pool of bilingual service representatives than was available to them at the individual field offices. 1 Unlike its changes to the UI program, the department’s addition of telephone assistance to the DI program simply complements the existing DI claim-filing and information process. Created in 1946, the DI program provides weekly benefits to workers who lose their wages generally due to pregnancies or to illnesses or injuries that are unrelated to work. Callers to the toll-free number that the department recently established for the DI program cannot file claims over the telephone. Instead, the number gives callers access to general information about the program and its services seven days a week for a minimum of 15 hours each day. To assist callers who want to obtain informa- tion from customer service representatives, the department established two call centers as well as customer service units in 5 of its 17 field offices. Customers of both programs are generally satisfied with the customer services they have received over the telephone. In several surveys of customers that the department conducted for the UI program, at least 82 percent of survey participants indi- cated that the assistance they received was prompt, convenient, and reliable. Nevertheless, the customers also voiced concerns about lengthy wait times and the customers’ inability to reach a service representative. Surveys of callers to the DI call centers found that customers were similarly satisfied with the services supplied over the telephone. On the other hand, the customers also identified lengthy waits as their primary concern. The department plans to conduct additional surveys of the DI program this year. Although the department’s efforts have improved customer service and increased customer access to the programs, custom- ers have experienced difficulties when requesting customer assistance. Specifically, callers to the UI program’s toll-free telephone numbers have experienced lengthy wait times during certain busy periods. For example, more than 60 percent of the callers during a peak service period in February 2001 waited on hold five or more minutes. This percent represents a sharp increase from the 18 percent who waited five or more minutes during December 2000. Customers of the DI program experienced other problems. For instance, over a 15-month period from January 2000 through March 2001, the telephone system at DI call centers required nearly 687,000 (27 percent) of the 2.5 million callers who asked to speak to a customer service representative to call again. Moreover, nearly 31,000 callers routed to the DI program’s call centers in the first 3 months of 2 this year received busy signals, and this figure is considerably higher than the 850 callers who encountered busy signals during the same period last year. According to the department, the difficulties encountered by customers relate in part to staffing shortages at the call centers for both programs. The department attributes the shortages to various problems, including slow recruiting, a lengthy hiring process, and trouble retaining customer service representatives because they typically move into other, higher-paying positions. In addition to the hiring difficulties, the DI program has faced numerous phone system problems that have hampered callers’ ability to receive services when they desire them. Moreover, until March 2001, the department allowed relatively few callers for the DI program to wait on hold, thereby increasing the number of callers who had to call back. To improve its customer service for the UI and DI programs, the department established goals for its desired response time to customers. For the UI program, the department’s goal was to answer within four minutes 90 percent of all calls requesting information and to answer in five minutes 90 percent of all claim-filing calls. However, the department cannot measure whether it has met its goals because one of its system modifica- tions eliminated its ability to distinguish information calls from claim-filing calls. Nevertheless, our review of monthly call data after the department’s implementation of the toll-free telephone numbers revealed that in 8 of the 16 months from November 1999 through February 2001, at least 10 percent of callers for the UI program waited five or more minutes on hold to speak to a customer representative. During the 16 months, the percent of callers waiting at least five minutes ranged from 3 percent to 61 percent. UI program management recognizes as a problem its inability to measure fully its performance and is currently evaluating a proposed goal that it can use to measure the response time of its customer service system. For the DI program, the department set a goal for the call centers and customer service units to answer in four minutes 90 percent of all calls requesting information. However, it evaluates the program’s performance from management reports that do not routinely include the customer service units, which receive 42 percent of the program’s calls. Addition- ally, the management reports do not indicate its performance in meeting its stated goal. Nevertheless, our analysis of the calls to 3 the DI program’s two call centers for a 15-month period from January 2000 through March 2001 disclosed that 90 percent of the callers who waited to speak to a customer service representa- tive remained on hold for less than five minutes. RECOMMENDATIONS To continue enhancing customer service for its UI program, the department should take these steps: • Continue to explore ways within the State’s civil service system to hire part-time or seasonal staff to reduce customer wait times during expected busy periods. Additionally, the department should reassess the minimum qualifications and use continuous testing for its customer service representa- tives. However, before hiring a large number of staff, the department should perform a cost-benefit analysis to deter- mine the number of staff it should hire. • Consider performing a study to examine the effect on call center workloads of increasing business hours for call centers during peak calling periods. • Promptly complete its process for setting challenging yet reasonable goals for answering customer calls. It should also modify its management reports to measure its performance in meeting these goals. To ensure that customers for the DI program continue to have access to telephone services, the department should do the following: • Continue to explore and use the various methods allowed in the State’s civil service system to hire promptly and then retain call center staff. • Complete contingency plans for customer service and work closely with its telecommunication contractors to limit the effect and number of system breakdowns during installation of future system changes. • Modify its management reports to include the call activity at its customer service units and call centers and to measure and report its performance in meeting its goal. 4 AGENCY COMMENTS The department generally concurs with our findings and is n taking steps to implement our recommendations. 5 Blank page inserted for reproduction purposes only. 6 INTRODUCTION BACKGROUND D uring the 1990s the Employment Development Depart- ment (department) began providing information and services over the telephone to customers of its unemployment insurance (UI) and disability insurance (DI) programs. The UI program shifted its main method of providing services—primarily filing claims for unemployment benefits and disseminating information—from supplying assistance in person to providing help over the telephone. In contrast, the DI program merely supplemented with telephone services its existing customer assistance methods, and its telephone services do not allow callers to file claims over the telephone. However, the Legislature raised concerns about the promptness of the services that the programs provided to customers. As a result, Chapter 329, Statutes of 1998, required the department to address specific concerns by enhancing its customer services for the programs. In particular, the statute called for the department to establish toll-free telephone numbers for customers of each program and, to the extent mandated by state law, to furnish telephone assistance in languages other than English. The statute further required the department to ensure customer access to services in person if requested, to provide in writing accurate and timely claim information to customers, and to furnish requested case information to their customers in a timely manner. The UI Program Shifted Its Focus From Providing Services in Person to Providing Services by Telephone Until the mid-1990s the department required customers to visit local field offices to file unem- Unemployment insurance, established ployment claims. Customers filing claims were in California in 1938, is intended to partially compensate individuals for lost further required to supply the department with wages. To be eligible for benefits, documentation to establish their eligibility. The individuals must be unemployed through department estimated that during busy periods in no fault of their own, able and available its field offices, customers had to spend at least a to accept full-time work, and actively few hours in order to file claims. Moreover, when seeking work. The department generally customers did not submit all the necessary provides benefits for up to 26 weeks. information to file claims, they had to return on another day. 7 To address its fluctuating workload and improve customer access to the UI program, the department began in 1994 and 1995 a series of pilot projects for telephone claim filing. Surveys of UI customers in rural and metropolitan areas in California indi- cated customers’ satisfaction with telephone claim filing, so in October 1996 the department implemented these services throughout the State. With the original telephone services, UI customers called a local or long-distance telephone number to file their claims. The telephone system then routed the calls to the nearest of several call centers that the department had established to service its UI customers. UI Customers Now Use Toll-Free Telephone Numbers to Access Services In September 1999, responding to the Legislature’s concerns, the department fully implemented its toll-free telephone system for claim filing and for requesting information for the UI program. To do so, the department established four 800 numbers for its customers who speak English, Spanish, Cantonese, and Vietnamese. It had already established toll-free numbers for its deaf and hearing-impaired customers in September 1997 and for its out-of-state customers in October 1996. Previously, individuals who had worked in California but who resided in other states filed claims at local unemployment offices in their state of residence and the docu- mentation was forwarded to the department for processing. Customers of the UI program file claims or gather information about the program by calling one of the toll-free numbers. To make the best available use of its customer service representa- tives in the six call centers it established in various locations throughout the State, the department uses intelligent call routing, which links the representatives as if they were in one location. Thus, regardless of the customer’s location, the routing system transfers a customer’s call to the call center with the greatest number of available representatives. Telephone Services Supplement the DI Program’s Claim Process In contrast to telephone services at the UI program, those for DI program customers supplement—and do not replace—the program’s existing claim filing and information process. Unlike callers to the UI program, callers to the DI program cannot file claims over the telephone. Instead, they can obtain information 8 in English and Spanish about the program and its benefits. Additionally, callers can obtain the claim forms they must complete when filing for DI program benefits. Until the department began providing telephone services, individuals obtained the forms solely from physicians, medical facilities, employers, and local DI field offices. The disability insurance program, Customers seeking information about the program created in 1946, provides weekly benefits or their benefits usually visited field offices. to any worker who loses wages when the individual is unable to work for at least eight days generally due to an illness or The department established call centers in injury unrelated to work or because of Sacramento and Riverside in 1998 and 1999, pregnancy. A physician or medical respectively, to make services more accessible and practitioner must certify the disability. convenient for customers of the DI program. The department provides benefits for a Initially, customers called local telephone numbers maximum of 52 weeks. when they needed information. To improve access further, the department added customer service units at 5 of its 17 DI field offices in 1999 and 2000. In September 2000 the department established a toll-free telephone number for its DI program customers. The department’s telephone system allocates the toll-free calls among the call centers and the customer service units according to percentages of available staff. The department modifies the percentages periodically based on changes in the numbers of employees at the call centers and at the customer service units. Otherwise, the system used by customers when seeking telephone services for the DI program is similar to that of the UI program. The Call Centers’ Responses to Service Requests Depend on the Call Types and the Centers’ Levels of Activity As Figure 1 shows, callers to the UI and DI toll-free telephone numbers either enter the telephone network or receive a mes- sage that the network is busy and cannot answer the call. The message also tells customers to call back at another time. When callers enter, they reach an interactive voice response system (voice response system), which is a series of automated menus that describes the various options that customers can select depending on the nature of their calls. For example, callers to the UI program can obtain information about filing a claim, finding jobs, and locating field offices. In addition, callers for both programs can learn the status of their benefit checks or receive general information. 9 FIGURE 1 The Flow of Calls to the Employment Development Department’s Call Centers* Caller dials a toll-free telephone number Cal v le o r i c e e n r t e e s rs p o th n e s e in s t y e s r t a e c m tive No ne Is t w th o e r k Yes s C ig a n ll a e l r o re r c a e m ive e s s s a a g b e u s t y o busy? call back later Subject of call Status of benefit check Request to file a claim† General information Location of the Information about jobs† nearest office† and employers Customer asks to speak to a service representative Has the Are all system reached service Yes the established Yes representatives No limits for busy? call volume? No The system places the caller in a queue to wait on hold The system deflects Customer abandons call Representative and disconnects call (caller hangs up) answers call * This flow chart shows the call process for customers speaking English and Spanish. Calls to the toll-free numbers for Cantonese and Vietnamese customers bypass the interactive voice response system; otherwise, the process is the same. Calls to the toll-free teletypewriter numbers are answered immediately by customer service representatives, or the call waits in a queue. † Applies to the callers for the unemployment insurance program. When a caller wants to file a UI claim or opts to speak to a customer service representative to obtain information, the network routes the call to a call center or a customer service 10 unit. Depending on the availability of customer service representatives, the center or unit may respond to the call in one of three ways: • Answer the call immediately. • Deflect the call. • Place it in a waiting line, called a queue. The telephone network deflects a call when the system has received its maximum number of calls. A recorded message advises the caller of the optimum times to call back to avoid delays in obtaining services. The system then disconnects the call. If all customer service representatives in the call centers are busy but the limits have not been reached, the network places the calls in the appropriate call center’s queue. While waiting, callers receive recorded information about the programs, and callers for the UI program may hear about claim-filing require- ments. Customer service representatives then answer the calls eventually, or callers abandon the calls before representatives answer. Callers abandon—hang up—for various reasons. For example, they may not have the required information, or they may decide the wait is too long. Statutes Also Required the Department to Report on Staffing and on the Progress of Its Telephone Services Statutes in 1998 and 1999 directed the department to identify the resources necessary for its call centers and units to respond quickly to customers calling about the UI and DI programs. In addition to requiring customer service enhancements for both programs, Chapter 329, Statutes of 1998, required the department to report to the Legislature the appropriate staffing levels needed to answer all incoming calls so that no customer waited more than three minutes before talking to a service representative. In response, the department reported in January 1999 that it could not meet the three-minute goal. Specifically, it maintained that the cost to achieve sufficiently the three-minute goal for the UI program ranged from $53.5 million to $71.6 million, significantly more than its $36.2 million budget for that fiscal year. At the same time, the department estimated that it needed $10.1 million for the DI program, which was $1.5 million more than the $8.6 million budget it projected it would need after it had fully established its call centers. 11 Chapter 50, Statutes of 1999, further required the department to report to the Legislature about its progress in implementing toll-free telephone services for filing UI claims. For example, the statute required an estimate of the appropriate staffing levels needed to answer all calls within an average of three minutes. In its March 2000 report, the department estimated that it needed between 832 and 1,192 staff to answer within the time limit all the calls received by its UI call centers. To achieve the desired number of staff, the department reported that it would require between $55.7 million and $79.8 million. The department budgeted 681 staff for its UI call centers at a cost of $45.6 million for fiscal year 1999–2000. It also estimated that the UI program would need some of the additional staff to answer calls during peak calling periods only. SCOPE AND METHODOLOGY Chapter 329, Statutes of 1998, required the Bureau of State Audits to review the implementation of toll-free telephone claim services for the UI and DI programs. In particular, the statute required that we examine several factors that affect the department as well as UI and DI program customers. The Appendix identifies the factors we reviewed and briefly describes how the implementation of telephone services affected the department and the customers of its UI and DI programs. To analyze the impact on customer service of the toll-free tele- phone system, we identified the customer service goals that the department established for both programs. We obtained and reviewed monthly call volume reports for the UI program for November 1999 through February 2001. Additionally, we compared call volume on the first day of the workweek with that for other days over nine weeks from October through December 2000. Moreover, we reviewed staffing forecasts and contrasted the forecasts with actual staff at the call centers. We also examined information about the numbers of authorized positions and vacancies at the UI call centers. Further, we analyzed data about the idle time of call center staff from March 2000 through March 2001. For the DI program, we reviewed monthly call volume reports for its two call centers from January 2000 through March 2001. Additionally, we analyzed information about the number of authorized positions in the call centers and the customer 12 service units as well as their vacancies as of April 2001. We also evaluated data about the idle time that call center staff experi- enced from January 2000 through March 2001. To gauge customer satisfaction with telephone services, we obtained customer service surveys that the department’s UI and audit and evaluation divisions conducted for the UI program during the last several years. Additionally, we obtained the surveys of callers to the DI call centers that the audit and evaluation division performed in 1998 and 1999. We evaluated the content and structure of all the surveys and ensured their design yielded results upon which we could rely. Using the results of the surveys, we assessed the level of customer satisfac- tion with the telephone services provided by the department for both programs. Further, we identified those aspects of the service for which customers of the UI and DI programs expressed their concerns. We assessed the impact of telephone services on customers’ ability to access routine and nonroutine information about individual cases. Specifically, we ascertained the types of program and claim-filing information offered in the voice response systems when customers call the toll-free telephone numbers. We also identified the types of information included in individual customer records in the databases used for the programs. Additionally, we determined how callers learn about access to services provided in person in local field offices. In evaluating the effect of telephone services on the number of claimant appeals filed, we obtained data about claims and appeals filed by customers of the UI and DI programs. We analyzed for the UI program claim and appeal data for December 1997 through December 2000. For the DI program, we reviewed similar data for the same period. Furthermore, we assessed for the UI and DI programs whether other factors affected changes in claims and appeals filed. We found that the implementation of telephone services has had no noticeable impact on the number of claimant appeals filed in the UI and DI programs. Because of legislative concerns about the costs associated with the department’s having numerous providers of its telephone services, we obtained from the department its UI and DI expenditures before and after it implemented the toll-free numbers. We reviewed and analyzed the expenditure data to ascertain the changes in costs since the advent of its telephone assistance system. However, our review disclosed that when the 13 department established its call centers and toll-free numbers, it reduced the number of telephone service providers it used throughout the State. Thus, the department has addressed the concerns about having numerous providers of telephone ser- vices. Consequently, we performed no further review. To evaluate the effect of telephone services on UI and DI customers’ ability to furnish written documentation to the department, we determined when the department requires customers to submit forms and other written materials necessary for filing claims or for supplying information. We also identified the methods that customers use to submit information in writing to the department. We further verified how the department ensures that staff includes accurately in a customer’s case file the submitted information. Moreover, we ascertained the procedures used when callers request a copy of their claim record. Finally, we determined whether the department mails notifications or requested information within a specified number of days to ensure that UI customers receive these materials in a timely manner. Our audit also evaluated the impact of telephone services on customers with disabilities. We reviewed the Americans with Disabilities Act (act) to identify the level of services that the act requires the department to provide to its disabled clients. We then gained an understanding of the types of disabilities most often affected by telephone services. Based on our review of the act and our understanding, we focused our review on callers with speech and hearing impairments. We then obtained and reviewed data on the calls received from the toll-free teletype- writer numbers that the department established for customers of each program. Finally, we ascertained whether the implementation of telephone services affected the language services that the department provides to its UI and DI callers. We identified the threshold for providing services in languages other than English mandated by the Dymally-Alatorre Bilingual Services Act. We also reviewed the department’s policy for providing services in other languages. Further, we reviewed the department’s 1996 and 2000 surveys of bilingual services required by the act. We pinpointed the languages that participants in the surveys spoke that met the mandated threshold, and we determined whether the department made any changes in the language services it offers its customers. Moreover, we identified the other languages for which the department provides interpreters to n UI and DI callers. 14 CHAPTER 1 Unemployment Insurance Customers Receive Improved Services Over the Telephone but May Endure Long Wait Times During Peak Calling Periods CHAPTER SUMMARY T he implementation of telephone services by the Employ- ment Development Department (department) for its unemployment insurance (UI) program has improved the program’s claim-filing process and made information readily accessible to customers. However, during certain peak calling periods, customers must endure lengthy wait times before speaking to a customer service representative. On the other hand, the department estimated in 1995 that service provided in person by staff in an urban field office required as much as 4 hours and 32 minutes of an applicant’s time. After developing its telephone assistance system, the department estimated that a person could receive help in as little as 15 minutes. Our analysis revealed that the average call length—for claim filing and for information inquiries from April 2000 through March 2001—is approximately 9 minutes. Since it introduced statewide telephone claim-filing in October 1996, the department has made changes that not only improve its ability to assist customers but that also use the department’s resources more efficiently. With intelligent call routing, which links the six call centers for the UI program that are located throughout the State, program employees spend little time idle, and the department can more efficiently serve its customers with fewer employees. Furthermore, by linking call center service representatives, a larger pool of bilingual staff is available to respond to customers. Currently, the department provides a toll-free number to its Spanish-speaking customers, who represent 19 percent of its callers. The department also has toll-free telephone numbers for Cantonese- and Vietnamese- speaking callers, and language services are available for 25 additional languages. The department believes that people speaking these languages now receive more consistent service over the telephone than they did when they had to obtain services in person. 15 Moreover, the various surveys that the department conducted over the last five years show that customers are generally satis- fied with the services they receive over the telephone. In an early survey, conducted during the department’s pilot of telephone claim filing, customers rated telephone services as better than services provided in person. In subsequent surveys, at least 82 percent of the callers surveyed indicated that the service they received was prompt, convenient, and reliable. Nevertheless, some surveyed customers voiced concerns about lengthy wait times and about their inability to reach a customer service representative each time they called. Despite the enhanced customer service that the department’s telephone system offers, the system requires further improve- ments to shorten customer wait times during certain peak periods. Our analysis showed that customers are more likely to wait longer during the program’s busiest months— January through March—or on the first day of the workweek. The department asserts, and we agree, that it does not make good business sense to hire a significant number of permanent employees to avoid long customer wait times during peak periods because these customer service representatives will have idle time during non-peak periods. However, the department should explore ways within the civil service system to hire part-time or seasonal staff to reduce customer wait times during expected busy periods. It should also consider expanding the business hours that representatives are available during these busy periods to help accommodate the increased workload. To serve customers of the UI program, the department established goals to answer 90 percent of information calls within four minutes and 90 percent of claim-filing calls within five minutes. However, the department has been unable since September 1999 to measure whether it has met its goals because one change to the telephone system eliminated the department’s ability to distinguish information calls from claim-filing calls. Our review of monthly call data since implementation of the toll-free telephone numbers revealed that in 8 of the 16 months from November 1999 through February 2001, at least 10 percent of UI program customers waited five minutes or more for customer service representatives. Over the 16 months, the percent of callers waiting at least five minutes to speak to a customer service representative ranged from 3 percent to 61 percent. The department recognizes as a problem its inability to measure fully its performance, and it is currently developing a new, measurable performance goal. 16 THE ADVENT OF THE DEPARTMENT’S ASSISTANCE BY TELEPHONE HAS IMPROVED CUSTOMER SERVICE FOR THE UNEMPLOYMENT INSURANCE PROGRAM The introduction of telephone assistance for the UI program has enhanced customer service and made information readily available to customers. By reducing the time required to file a claim, offering services in multiple languages, and providing teletypewriter service to the speech and hearing impaired, the department has improved customer satisfaction. Additionally, customers can more easily obtain information through the department’s interactive voice response system (voice response system), which provides general information 24 hours a day. Customers Save Time by Using the Program’s Telephone Services Since the department introduced telephone assistance, the tasks of filing for unemployment insurance and asking questions about unemployment insurance have become less burdensome to customers. Previously, customers filed unemployment claims in person, and this method took customers significantly more time than filing claims over the telephone. Currently, UI customers call a toll-free number to file claims and to obtain information. In October and November 1995, during its pilot of telephone services, the department conducted a survey to show that filing claims by telephone takes less time than filing claims in person. The department surveyed one urban area and one rural area. According to the survey, the department estimated that custom- ers in the urban area spent an average of 2 hours and 48 minutes of their time filing a claim in person. The filing time increased to 4 hours and 32 minutes if the customers were required to return for a follow-up visit. The department estimated that rural customers spent an average of 3 hours and 28 minutes filing a claim in person and 6 hours and 28 minutes if they had a follow-up visit. These estimates included measures of commute According to a time to and from a local job office and the time customers spent department survey, filling out forms. customers spend significantly less time According to the survey, customers spend significantly less time filing unemployment obtaining services by telephone than they had spent filing insurance claims by claims in person. Most of the time-saving resulted from the telephone than in person. department’s eliminating travel time for customers. However, customers also saved time because they had shorter waiting 17 times and because the customers no longer needed to fill out paper forms. The department’s estimates indicate that to complete the telephone filing, a customer typically spends From April 2000 through 15 minutes—4 minutes on hold and 11 minutes filing the March 2001, the average claim—for the entire process. call length after a customer requested We conducted a limited analysis of information and claim-filing assistance was nearly calls answered by customer service representatives. Our analysis nine minutes. revealed that from April 2000 through March 2001, the average call length after a customer requested a customer service representative was nearly nine minutes—three minutes on hold and six minutes speaking to a representative. The Department Offers Telephone Assistance in Multiple Languages The department offers customers translation services in various languages. It conducts a biennial language survey, as required by the Dymally-Alatorre Bilingual Services Act, to determine the number and percentage of customers who speak other languages. This act requires state agencies to provide adequate bilingual services for customers who speak a specific language if the number of customer contacts in that language equals or exceeds 5 percent of the people served by the agency. In its most recent language survey, which was conducted in January and February 2000, the department identified Spanish as the only language other than English that meets the requirements of the act, reaching 19 percent of the program’s call center contacts during the survey period. The department provides a toll-free number for its Spanish- speaking customers. Additionally, the department elected to provide toll-free numbers in Cantonese and Vietnamese. Moreover, in its call centers, the department provides translation services in 25 other languages, including Arabic, Farsi, Russian, and Cambodian. Because the six call centers are linked, all bilingual customer service representatives are in effect consolidated into one large pool. Thus, customers who require language services are no longer limited to the staff at the local job offices. Instead, bilingual staff in any one of the six call centers across the State are now available to customers who may need the staff’s lan- guage skills. The department asserts that the pooling of bilingual customer service representatives ensures a more consistent level of service to its non-English-speaking customers than did its former methods for serving customers at the local job offices. 18 The Telephone System Accommodates Customers With Speech or Hearing Impairments The department’s switch to customer service by telephone also accommodates its customers with hearing and speech impair- ments. The Americans with Disabilities Act requires that when a public entity communicates by telephone with applicants and beneficiaries, the entity must use telecommunication devices or equally effective telecommunication systems to communicate with individuals who have impaired hearing and speech. The department set up a separate toll-free number specifically for teletypewriter (TTY) calls. During our review of seven selected months between November 1999 to January 2001, TTY calls represented less than 1 percent of the calls made to the call centers. Callers who use the toll-free TTY number are able to connect directly to a customer service representative with the use of TTY technology. If the customer chooses not to use the direct toll-free TTY number, he or she may choose to use a relay operator and to call one of the voice lines. Surveys Show That Unemployment Insurance Customers Are Generally Satisfied With the Service They Receive Since the department began providing services by telephone in 1994, it has conducted several surveys to learn whether custom- ers are satisfied. During the pilot of its telephone services, surveyed customers preferred filing claims by telephone to filing claims in person at local job offices. Overall, surveys revealed general customer satisfaction. The customers rated the service based on various factors, including promptness, convenience, Unemployment insurance and reliability. Each category received a rating of 82 percent or customers rated their more regardless of the survey year. Customers rated “overall “overall satisfaction” satisfaction” as good no less than 78 percent of the time and as as good between high as 94 percent of the time. 78 percent and 94 percent of the time. In an earlier survey, customers stated that improvement in system components, such as the introduction of toll-free num- bers, would increase satisfaction. Since the early surveys, the department has implemented the toll-free numbers. When customers call a local job office for unemployment insurance service, the system no longer forwards the calls to a call center. Rather, a recording tells customers to call one of the department’s toll-free numbers. Thus, callers no longer pay toll charges. The surveys, conducted during the department’s pilot offering of telephone claim filing, revealed that customers are happier with the convenience of telephone services than they were with the 19 services offered in local job offices. Nevertheless, in subsequent surveys, customers voiced concerns about lengthy wait times or Customers voiced their inability to reach a customer service representative at concerns about lengthy certain times. Some customers suggested that the department wait times or their hire more service representatives. Despite these complaints, inability to reach a most customers stated in the survey that they were willing customer service to wait up to 10 minutes on hold to speak to a customer representative at service representative. We found that from November 1999 certain times. through February 2001, more than 1.7 million (16 percent) of 10.4 million customers asked to speak to a representative and waited on hold 5 or more minutes. Of the 1.7 million callers, over 318,000 (19 percent) waited 10 or more minutes. Prior surveys performed by the department did not isolate samples of specific customer groups, such as Cantonese or Vietnamese customers or TTY users. The surveys were not likely to get representation from these groups because their popula- tions are relatively small in relation to total claims filed. How- ever, the department has recently conducted surveys of these customer groups to also assess their satisfaction with services. It did not provide us with the results of these surveys because the sample sizes were too small to provide statistically valid results. The department plans to conduct additional surveys of these customer groups in the future. The Telephone System Can Answer Many Questions Day or Night Telephone service has allowed the department to expand its service hours. The department’s previous customer service methods limited a customer to an assigned appointment time during local job office hours. The department rigidly assigned appointments based upon customers’ social security numbers and the availability of its staff. Currently, customers can obtain information around the clock via the voice response system that the department installed. When customers call the English, Spanish, or out-of-state toll-free numbers, they may choose from a list of options that provides service and answers their questions. Callers can receive general information on topics including appeals, claim-filing requirements, eligibility interviews, employment recruitment and referral services, and job service locations. Customers can also hear recorded information in both English and Spanish about their most recent benefit checks, including the period covered and the date the checks were issued. This 20 information is available from 6 a.m. to 12 a.m. six days a week and from 6 a.m. to 9 p.m. on Sunday. Before implementation, the department did not provide check information to customers until 10 days after the customers mailed in their biweekly claim forms. If the voice response system cannot answer the customers’ questions or if customers prefer to speak to a person, they may choose the option to speak directly to a customer service representative at any time during their call when calling during regular business hours. Customer service representatives can provide callers with information on UI claims, help resolve eligibility issues, provide information on filing an appeal, or give general information on other department pro- grams and services. The Department Changed Its Telephone System to Reduce Customer Wait Times The department has continued to modify its telephone system to better use resources for handling incoming calls while still providing good customer service. In 1999 the department put into service a new system for routing customer calls. The system, The department’s called intelligent call routing (call routing), routes to the next implementation of available representative each call from a customer who asks to intelligent call routing speak to a representative. The representative who answers the has improved the call may be responding from any of the department’s six call productivity of its centers located throughout the State. Before call routing, the customer service department’s system routed calls to call centers based upon representatives. geographic service areas. At times, calls overwhelmed some call centers, and this situation caused customers to experience extended wait times. On the other hand, other call centers had representatives with idle time. The implementation of call routing improved the productivity of all customer service representatives. Our analysis of representatives’ idle time from March 11, 2000, to March 11, 2001, shows that on average, representatives were idle for only 24 seconds between calls. Also, call routing enhances the efficiency of customer service. For example, the department stated in its March 2000 report to the Legislature that the UI program needs about 534 representa- tives to assist 1,400 callers in a one half-hour period if each of the six call centers takes 233 calls. However, the department 21 maintained that with call routing, it needs only 514 representa- tives to assist the same number of callers when all six centers are linked as a single call center.1 CUSTOMERS SOMETIMES HAVE LENGTHY WAITS DURING PEAK CALLING PERIODS Although the average time a customer spent on hold between April 2000 and March 2001 was only 3 minutes, many individual callers waited beyond 5 and 10 minutes during this same period. Increased call volumes on the first day of the workweek or during the UI program’s busy season— January through March—caused many customers to wait 5 minutes or more to speak to a representative. Until the department can fill its current vacant customer service positions and have the appropriate number of staff available during peak periods, customers calling during these busy periods will continue to experience long wait times. Fluctuations in Call Volume Contribute to Long Waits Our analysis of call center data revealed that callers are more likely to encounter long waits during certain busy periods. Annually, call volume is typically highest between January and Callers are more likely to March and weekly on the first day of the workweek. Average encounter lengthy waits hold times may be low overall; however, individually and during peak calling during certain periods, wait times vary greatly. For example, the periods—January monthly average amount of time that a customer waited on through March and hold between April 2000 and March 2001 was approximately weekly on the first day of 3 minutes. However, many customers waited on hold much the workweek. longer than that. During this same period, a customer’s time on hold ranged from as little as none to as high as 67 minutes. The department explained to us that January through March is a busy period for the UI program because of high levels of unemployment in the retail industry after the holidays and in the construction and farming industry because bad weather is typical during those months. We found that, in general, more callers waited five or more minutes during these busy months than they did during the preceding months. Our analysis also showed that more customers cannot access or navigate the system because telephone lines are congested during this time. 1These staffing scenarios that the department reported to the Legislature assume that it takes 10 minutes to serve each customer and that no customer could wait more than 3 minutes for service. 22 Figure 2 confirms that during the first two or three months of the year, the department received more calls to the UI program’s toll-free numbers than in the preceding months. The figure also shows the number of callers that waited five or more minutes to speak with a customer service representative. FIGURE 2 The Number of Calls From UI Customers Tends to Increase in the First Few Months of the Year Source: Monthly call center reports from the Employment Development Department’s unemployment insurance program. *Calls include those from customers who use only the interactive voice response system, customers who receive messages to call back later, and customers who ask to speak to a customer service representative. 23 *sllaC fo rebmuN Total calls accessing the system Total calls accessing the system in one of the first few months of the year 1,750,000 Calls waiting five or more minutes to speak to a customer service representative 1,500,000 1,250,000 1,000,000 750,000 500,000 250,000 0 N ov e m b e D r ec e m b er J a n u ar y F e br u ar y M arc h A pril M a y J u n e J ul y A u g u S s e t pt e m b er Oct o b N er ov e m b e D r ec e m b er J a n u ar y F e br u ar y 1999 2000 2001 As anticipated, the busy months of January and February 2001 had a significantly greater percentage of callers waiting on hold for five or more minutes when compared to the other months of the year. Specifically, more than 40 percent of the callers who wanted to speak to a customer service representative waited on hold for at least five minutes in January. Although fewer customers called in February than in January, over 60 percent of February’s callers waited five or more minutes. According to the department, this is likely due to fewer workdays in the month and because there was a change in the calculation of customer In February 2001 over benefits, which usually increases the length of time a call takes. 60 percent of the callers In contrast, only 18 percent of the callers in December 2000 who wanted to speak to waited at least the same amount of time or more. a customer service representative waited on The first day of the workweek following a weekend or a holiday hold five or more minutes. is also typically a busy time for the UI call centers. The depart- ment stated that it receives about 33 percent of its total weekly UI calls on Mondays. Our analysis of nine consecutive weeks starting on October 1, 2000, and ending December 2, 2000, confirmed that over the nine weeks, 30 percent of the callers placed their calls on Monday or the first day of the workweek. Because of this influx of calls, more callers waited on hold on Mondays than on any other day of the week. In fact, our analysis shows that about 55 percent of the weekly calls where callers were required to wait five or more minutes occurred on Mondays. As shown in Figure 3, the percent of callers that had to wait five or more minutes is consistently higher on the first workday of the week. Certain Barriers Can Limit the Department’s Capacity to Provide Prompt Customer Service Despite the department’s intent to answer calls efficiently, several barriers prevent the department from always providing short wait times. Unpredictable fluctuations in call volume hamper the department’s ability to estimate its staffing needs correctly. Additionally, increased call volume during peak peri- ods often requires more staff than the department has available. The department maintains that it does not have the necessary federal funding to hire staff to answer all calls within three minutes during peak periods without an augmentation of state funds. The department also states that its workloads during peak periods have increased faster than it has been able to hire and train the additional employees needed. 24 FIGURE 3 Most Customers Who Wait Five or More Minutes Place Calls on the First Day of the Workweek Weeks Beginning (2000) Source: Weekly call center reports from the Employment Development Department’s unemployment insurance program. Forecasting the number of staff required to answer calls effi- ciently is essential to providing good service. The department uses a formula in its call management software called Erlang C, which is widely accepted for determining the staffing needs of call centers, to calculate the number of service representatives needed to answer calls. The software uses projected incoming call volume, among other factors, to calculate staffing levels. The department is limited in how well it can forecast the projected call volume because of call fluctuations due to uncontrollable events such as unusual weather conditions, natural disasters, and economic downturns. Also, call volume can fluctuate greatly throughout the day. As a result of these unexpected workflow patterns, the department’s calculations of staffing needs sometimes miss the mark. If the department miscalculates and understaffs its call centers, callers will encounter longer 25 sremotsuC fo tnecreP setuniM eroM ro eviF gnitiaW Other workdays First day of the workweek 100 80 60 40 20 0 1 8 5 2 9 5 2 9 6 b er b er er 1 er 2 er 2 b er er 1 er 1 er 2 o o b b b m b b b O ct O ct O ct o O ct o O ct o N o v e o v e m o v e m o v e m N N N waits. If it overstaffs its centers, customer service representatives have idle time and the department increases its costs of operation. Because service times depend upon fluctuations in workflow patterns and seasonal changes, it is difficult for the department to always answer calls as quickly as it would like. For example, The department cannot the department projected it needed as many as 879 customer always answer calls as service representatives on one day in January 2001 and only quickly as it would like 198 on a day in December 2000. If it were to employ the larger because its service times number of staff, many employees would be idle during slower depend on fluctuations in call periods. The department points out that it is not a good workflow patterns and business practice to hire full-time permanent staff to accommo- seasonal changes. date predictable peak service periods. Although we agree with this, the department could increase staff during peak periods with seasonal or part-time employees. The department contends that during a favorable economy it is difficult to attract employees who would work only seasonal or part-time hours. Furthermore, the department already had experienced difficulties recruiting for its current vacant positions. In fact, as of April 2001, it had 87 of its 649 autho- rized call center positions vacant. The department attributes its staffing shortages to the normal high turnover rate experienced in the call center industry and to the complexities and slowness of the civil service hiring process. The department states that if it were to employ the 1,192 customer service representatives that it estimates are needed during peak periods to answer all calls within three minutes, it would exceed its current federal funding level. Federal funding is based on the number of claims unemployed individuals file. Given the current favorable unemployment rate, new state funds would need to be allocated to the department to pay for additional staff, equipment, and training required above current funding levels. Nonetheless, the department must continue to be aggressive in its recruiting efforts to fill current vacancies in its customer service representative positions. These efforts may require the department to use continuous testing for service representative positions and to reevaluate the minimum qualifications of this position, which currently require candidates entering the workforce to have a four-year college degree.2 Also, the 2Candidates with prior work experience must meet one of three requirements. They must have 60 semester units of college and worked for two years providing services or information to the public, worked in the department for one year as a program assistant or program technician, or worked in the department for four years. 26 department could focus on segments of the population that are interested in part-time work, such as college students or retired persons. The department could even consider extending its normal work hours during predictable peak periods and effectively marketing this change to customers on its automated system to help spread and accommodate the high volume of calls during these periods. THE DEPARTMENT CANNOT ENSURE THAT IT HAS MET THE CUSTOMER SERVICE GOALS IT ESTABLISHED FOR THE PROGRAM The department set goals to minimize the amount of time that customers wait to speak to a customer service representative for the UI program. However, due to changes it made to the tele- phone system in September 1999, the department has been unable to measure whether it is achieving its goals. As a result, management cannot be sure that the department is meeting its objective for responding to a high call volume within a limited time period. The Department Established Goals to Minimize Wait Times for Customers The department has stated that the purpose of providing tele- phone assistance is so that individuals with varying needs can quickly and easily obtain access to services. The department’s objective is to answer all calls quickly and efficiently. To meet this objective, the department established goals to answer 90 percent of claim-filing calls within five minutes and 90 percent of information calls within four minutes. Departmental work groups created these goals based on experience from early pilot programs before the department implemented the toll-free numbers statewide. The Department Cannot Determine Whether It Has Met Its When the department Goals for Assisting UI Customers implemented intelligent call routing in September 1999, When the department implemented intelligent call routing it lost the capability to in September 1999, it lost its ability to measure achievement of measure whether it is its customer service goals. With intelligent call routing, different achieving the goals it call centers no longer answer either information calls or established for the claim-filing calls; rather, all call centers answer both types of calls. The system often transfers customer calls from an UI program. information queue to a claim-filing queue to minimize wait 27 times. Although beneficial for callers, the modification has made it difficult to determine reliably the number of calls in either group. Moreover, the computer system does not track the indi- vidual hold times of the different call types. As a result, the department has lost the capability to measure whether or not it is achieving its goals. The department recognizes this deficiency The department cannot as a problem and is currently evaluating an option to change its measure whether it is customer service goal to answering 85 percent of all calls in four achieving its goals minutes or less, which can be tracked by its system. because its computer system does not track the We also found that reports prepared for management do individual hold times of not detail how well the call centers are doing as far as meeting the different call types. the department’s goal. Management reports are informative; however, none addresses the goal specifically. The reports do not show the number of calls answered in a four-minute interval even though the telephone system is capable of reporting the information at that interval. Instead, the department cur- rently prepares its management reports showing intervals within 2, 5, and 10 minutes. Equally important, the reports cannot show the number of calls for information and the number of calls for claim filing. Consequently, the department cannot effectively manage the UI program and ensure that its perfor- mance goals are met when it does not require adequate tracking and reporting. FOR HALF THE PERIOD WE REVIEWED, A SIGNIFICANT PERCENTAGE OF CALLERS WAITED FIVE OR MORE MINUTES FOR UNEMPLOYMENT INSURANCE REPRESENTATIVES Our analysis of the department’s monthly data for the UI program’s call centers revealed that 10 percent or more of callers waited five or more minutes to speak to a customer service representative during 8 of the 16 months we reviewed. We analyzed call center data from November 1999 through February 2001 to determine whether the department was answering 90 percent of all offered calls within five minutes or less. An offered call is a call during which the customer asks to speak to a customer service representative. As Figure 4 shows, during 8 of the 16 months we reviewed, 10 percent or more of the callers asking to speak with a customer service representative waited five or more minutes. During the 16-month period, the percent of callers waiting at least five minutes to speak to a customer service representative 28 ranged from 3 percent in April 2000 to 61 percent in February 2001. In January and February 2001, a much larger percentage of callers waited than in previous months because the unemployment insurance call centers experienced a heavier-than-expected call volume. FIGURE 4 Percent of Callers Waiting Five or More Minutes to Speak With a Customer Service Representative for the UI Program 70 60 50 40 30 20 10 0 N ove m be D r ece m ber Ja n uary Fe br uary Marc h A pril May J u ne J uly A u g us S t e pte m ber Oct o be N r ove m ber Dece m ber Ja n uary Fe br uary Source: Monthly call center reports from the Employment Development Department’s unemployment insurance program. RECOMMENDATIONS To continue to enhance the customer service improvements it has made to its UI program, the department should take these actions: • Continue to explore ways within the State’s civil service system to hire part-time or seasonal staff to reduce customer wait times during expected busy periods. Additionally, the 29 gnitiaW srellaC llA fo tnecreP setuniM eroM ro eviF Less than 10 percent of callers waited five or more minutes 10 percent or more of callers waited five or more minutes 1999 2000 2001 department should reassess the minimum qualifications and use continuous testing for its customer service representa- tives. However, before hiring a large number of staff, it should perform a cost-benefit analysis to determine the number of staff it should realistically hire. The cost-benefit analysis should appropriately weigh the benefits of reduced customer wait times in peak periods against the increased program costs of expanding its call center facilities to house additional staff and increased personnel costs to hire, train, and maintain increased staffing levels. • Consider performing a study to examine the effect on call center workloads of increasing business hours for call centers during peak calling periods. • Promptly complete its process for setting challenging yet reasonable goals for answering customer calls. Additionally, it should modify its management reports to measure its performance in meeting these goals. • Conduct planned customer satisfaction surveys of TTY users n and Cantonese- and Vietnamese-speaking customers. 30 CHAPTER 2 Despite the Department’s Efforts to Improve Access to Program Information, Problems Hamper the Disability Insurance Program’s Customer Service CHAPTER SUMMARY T he Employment Development Department (department) has taken steps to improve customer service by implementing a toll-free telephone system that offers information to customers of the department’s disability insur- ance (DI) program. Department staff receive customer calls at two call centers and five customer service units at field offices. However, staffing shortages and system problems have reduced customer access to the new system. Our review disclosed that over a 15-month period from January 2000 to March 2001, 27 percent of the customers asking to speak to a representative at DI call centers heard messages that told them to call back later and then had their calls disconnected. Also, during the first three months of 2001, nearly 31,000 callers using the system received busy signals when the system routed their calls to a DI call center. At its two DI call centers, the department is having difficulties in hiring and retaining employees. As of April 2001, nearly 30 percent of authorized customer service representative posi- tions at the two call centers were vacant. In addition, another 13 percent of its authorized customer service representatives were in training and unavailable to answer customer calls. Even after these representatives complete training in July 2001, the department will have filled only 71 percent of the authorized call representative positions with individuals who are ready to answer calls. This staffing deficit has resulted in many callers’ encountering busy signals and in callers’ hearing messages to call back later. Additionally, recent system failures have hampered the DI program’s capacity to supply services to callers. In January 2001 the department installed new equipment in its toll-free tele- phone information system. However, in the first three months 31 after the system’s installation, 15 different incidents caused parts of the system to shut down. When the system is down, callers may experience a busy signal. For example, during the first three months of 2001, nearly 31,000 call center customers received busy signals when telephoning the DI program’s toll-free num- ber. In contrast, only 850 call center customers received busy signals during the same period the previous year. The depart- ment has resolved these problems and attributes the incidents to start-up problems related to installation of the new equipment. The department also does not ensure that it has met its customer service goal for the DI program. The department established a goal of answering 90 percent of all DI information calls in four minutes. However, for its customer service units, which receive 42 percent of the system’s calls, the system cannot track the number of calls or the wait times to determine how long staff take to answer the calls. Despite these staffing shortages and system problems, the department has taken numerous steps to improve customer service and to increase the accessibility of information about the DI program. In addition to offering the toll-free number for its callers who speak English or Spanish, the department established a toll-free telephone number to better serve the needs of its hearing-impaired callers. The interactive voice response system (voice response system) that the department installed now allows customer access to information seven days a week. Furthermore, to reduce the number of times the system tells customers to call back and the number of calls receiving busy signals, the department has adjusted the telephone system so that it allocates calls among the two call centers and the customer service units. STAFFING SHORTAGES AT CALL CENTERS REDUCE CUSTOMERS’ ACCESS TO DISABILITY INSURANCE SERVICES The DI program lacks sufficient staff to accommodate the volume of callers to the program’s toll-free telephone number. Staffing shortages in the Unlike the UI program, which has known peak periods of DI program have affected service, the DI staffing shortages have affected its customer customer service on a service on a consistent basis. Specifically, staffing shortages consistent basis. have resulted in callers wanting to speak to a customer service representative being frequently asked to call back later and then disconnected from the system. 32 The DI Call Centers Are Significantly Understaffed The DI program has a vacancy rate of nearly 30 percent at its two call centers in Sacramento and Riverside. As shown in the Table, the call centers had 165 authorized positions for customer service representatives as of April 2001. However, 48 (29 percent) of those positions were unfilled. TABLE DI Call Centers Had High Vacancies as of April 2001 Percent of Percent of Percent of Authorized Authorized Authorized Authorized Representative Representative Representatives Representative Representative Positions Representatives Positions Available Positions Call Center Positions Vacancies Vacant in Training in Training for Calls Available for Calls Sacramento 74 14 19 4 5 56 76 Riverside 91 34 37 17 19 40 44 Totals 165 48 29 21 13 96 58 Source: Personnel records from the Employment Development Department’s disability insurance program. Another 21 representatives (13 percent) are being trained; thus, they are not available to take calls. As a result, the DI program only had 58 percent of its authorized representative positions available to take calls in April 2001. Even after the representa- tives complete their training in July 2001, because of existing vacancies the DI program will still only have 71 percent of its authorized customer service representatives available to take calls. Unlike the call centers, the five customer service units, which have 62 representatives, were not experiencing any vacancies as of April 2001. Staffing Deficits Restrict Customers’ Access to Program Information With the DI program’s staffing shortages, callers may find it more difficult than usual to obtain information from the program’s toll-free telephone system. When determining the number of callers it allows to wait in the queue to speak to a representative, the automated telephone system factors in the 33 number of representatives available to take calls. In March 2001 the department set the queue size for the DI program at 1.5 callers for each available representative. Currently with this queue size and 96 available representatives at the call centers, a maximum of 144 callers can wait on hold to speak to a represen- tative at any one time. However, if the call centers were fully staffed, the system queue size would increase 72 percent, thus allowing 247 callers to wait. When the number of callers exceeds the maximum queue size, the telephone system deflects callers and asks them to call back From January 2000 later to receive service. Our analysis of the 15 months from through March 2001, the January 2000 through March 2001 showed the system deflected DI telephone system almost 687,000 (27 percent) of the 2.5 million callers trying to deflected almost 687,000 reach a representative and then asked these customers to call (27 percent) of the back. As Figure 5 indicates, the percent of deflected calls each 2.5 million callers trying month ranged from approximately 10 percent to 45 percent over to reach a representative. the 15-month period. Currently, the department has a DI work group assigned to develop a service level goal for deflected calls. The group expects to present this goal to executive management for approval in September 2001. The department claims that difficulties in retaining and hiring call center employees have caused its staffing shortages. First, the DI program is unable to retain easily its call center employees, a problem that is generally expected in the call center industry. For example, after completing the long and often difficult process of hiring individuals into state service, the department has seen a high number of its call center repre- sentatives transfer to other less stressful or higher-paying jobs within state service. Additionally, the department struggles with recruiting potential candidates for customer service representa- tive positions. The department currently requires a four-year college degree for candidates without work experience.3 Further, the department maintains that the lengthy civil service hiring process has caused potential call center candidates to accept positions with private sector firms before it can extend an offer of employment. Generally a candidate needs 12 weeks to go through the department’s hiring process and an additional 12 weeks of training. In April 2001 the department implemented a continuous testing period for customer service representatives that it hopes will speed up the hiring process and increase its labor pool. 3Candidates with prior work experience must meet one of three requirements. They must have completed 60 semester units in college and worked for two years providing services or information to the public, worked in the department for one year as a program assistant or program technician, or worked in the department for four years. 34 FIGURE 5 Call Centers Deflect a Significant Portion of Calls From DI Customers Who Ask to Speak to a Customer Service Representative 50 40 30 20 10 0 Ja n uar y F e br uar y Marc h A pril May J u n e J uly A u g ust S e pt e m b er Oct o b e N r ov e m b er D ec e m b er Ja n uar y F e br uar y Marc h Source: Weekly call center reports from the Employment Development Department’s disability insurance program. 35 sllaC detcelfeD fo tnecreP 2000 2001 SYSTEM PROBLEMS HAVE PREVENTED CALLERS FROM RECEIVING SERVICE In January 2001 the department installed new equipment in its telephone system. However, during the first three months after installation, the telephone system encountered 15 different incidents in which parts of the system were down for up to one day because of equipment failure. For example, on March 30 the Sacramento call center was not able to receive any customer calls for over half the business day. This call center typically takes 33 percent of the calls coming into the voice response system. During two consecutive days in February 2001, callers routed to the Sacramento call center and all five customer service units were unable to check the status of benefit checks on the voice response system because of system problems. The Sacramento call center and the customer service units generally receive 74 percent of the customer calls. According to a department telephone system manager, these system failures resulted from implementation problems after upgrading the system and installing new equipment. However, he maintains these prob- lems have been corrected. When parts of the voice response system are down, callers may receive a busy signal or be deflected and asked to call back later to speak to a customer service representative. As a result, during this three-month period, nearly 31,000 callers received busy signals when they tried to contact the department’s DI call centers. During this same period the previous year, only 850 callers received busy signals. By the end of 2001, the DI program will install a personal identification number (PIN) system in the voice response system to improve security of customer information. To minimize the implementation problems that occurred with the previous system upgrade, the DI program is making contingency plans this time to ensure that customer service is not disrupted. THE DISABILITY INSURANCE PROGRAM CANNOT FULLY MONITOR ITS PERFORMANCE IN MEETING ITS SERVICE LEVEL GOALS One of the department’s objectives for the DI program is to serve its customers in a timely manner. Thus, it set a customer service goal to answer 90 percent of all calls within four minutes. However, currently the DI program cannot track the number or wait times of the calls answered at its five customer service units, which receive about 42 percent of the program’s calls. Unfortu- nately, the sophisticated computer software that tracks call Currently, the DI program information at its two call centers cannot be installed in the cannot track the number existing telephone equipment of the customer service units. or wait times of calls Instead, the DI program periodically estimates figures for the answered at its five number of calls answered, abandoned, and deflected in the customer service units, customer service units based on actual call experience from its which receive about Riverside call center. Consequently, regular monthly reports to 42 percent of the management include only call center activities, which account program’s calls. for approximately 58 percent of the calls received. Further, the call centers’ monthly report provided to the department director reports only the number and percent of calls answered within intervals of 2, 5, and 10 minutes or more. Because the monthly report neither includes the calls received in the customer service units nor identifies a four-minute interval for the calls received in the call centers, the department cannot ensure that it is meeting its customer service goal for the DI program. 36 Nevertheless, our analysis of available monthly call center data showed that from January 2000 through March 2001, 90 percent of the callers who waited to speak to a customer service repre- sentative waited on average less than five minutes. Figure 6 shows the percent of calls answered or abandoned within five minutes during the 15-month period. FIGURE 6 Customers of DI Call Centers Generally Wait Less Than Five Minutes Ja n uary Fe br uary Marc h A pril May J u ne J uly A u g us S t e pte m ber Oct o be N r ove m ber Dece m ber Ja n uary Fe br uary Marc h Source: Monthly reports from the customer service centers for the Employment Development Department’s disability insurance program. Although a relatively small percent of the call centers’ customers waited more than five minutes during most of the months we reviewed, more customers waited at least five minutes from January to March 2001. The department attributes the increase in callers waiting more than five minutes to its pilot program to change the queue size and the system problems due to installation of new equipment into the telephone system. THE DI PROGRAM HAS TAKEN STEPS TO IMPROVE CUSTOMER SERVICE The DI program has implemented major changes in its service delivery system to make services more convenient and accessible for customers. It established a statewide toll-free number, which 37 sllaC fo tnecreP denodnabA ro derewsnA setuniM eviF nihtiW 100 80 60 40 20 0 2000 2001 now links the two call centers and five customer service units and allows the program to allocate calls based on the number of representatives at each of the centers or units. Before the statewide toll-free number, calls were routed to a call center or customer service unit based on the caller’s geographic location. As a result, callers often received busy signals if the call volume to a specific customer service unit or call center was high. Recent linking of the call centers and customer service units has helped reduced the number of busy signals except during periods of system failure. To enhance service, the DI program also established a toll-free teletypewriter (TTY) phone number. The TTY service allows The department’s speech and hearing-impaired customers to receive information installation of a voice services directly without having to use an interpreter service. To response system has improve the service to Spanish speakers, who represent more increased customers than 13 percent of the department’s customers, information is access to information. available in that language in the voice response system when customers call the toll-free number. The service enhancement complies with the Dymally-Alatorre Bilingual Services Act, which requires state agencies to provide services in a specific language when 5 percent of people the agency serves speak a language other than English. In addition, the DI has service representatives capable of communicating in 34 other languages, including Cantonese, Laotian, Punjabi, and Romanian. Furthermore, the voice response system the department installed in its statewide toll-free number has increased the accessibility of information. This system allows callers access to information 18 hours per day, Monday through Saturday, and 15 hours on Sunday. Callers can receive information about how to file a claim, order needed forms or a payment history, check the status of their claim or benefit check, and receive general claim and benefit information without speaking to a representative. If a caller needs to speak to a representative, the system can transfer the call to an available representative during business hours. In an effort to decrease the number of deflected calls requiring the caller to call back at a later time to receive service, the department recently adjusted the queue size to allow more callers to wait on hold. Specifically, after a pilot study of using different queue sizes, the department increased its queue size from 1 caller to 1.5 callers per available customer service representative in March 2001. However, this increase in queue 38 size will also cause longer wait times if the department does not increase its number of available customer service representatives because more callers will be allowed to wait in the queue. After implementation of the statewide toll-free number in September 2000, the DI program established a work group to examine the program’s customer service goals. The work group plans to determine reasonable goals and make suggestions about how to improve customer access to DI services. Specifically, the work group is charged with recommending new customer service goals for the percentage of calls that representatives should answer in an allotted time, and with establishing reason- able percentages for calls that the program abandons, deflects, and allows to receive busy signals. The department conducted in 1998 and 1999 customer satisfac- tion surveys of DI customers who contacted its call centers. The surveys showed that callers during this time were generally satisfied with the service they received. The surveys asked customers questions regarding the quality of the service they received and the difficulty contacting the department by telephone. The surveys found that at least 77 percent of the customers rated their overall experience as good or very good. Additionally, no less than 66 percent of the customers said it was easy to reach the department by telephone. However, 15 percent of the customers in the surveys found it difficult to reach the department by telephone and cited the primary reason was that they had to wait too long to speak with someone. The department plans to conduct further customer satisfaction surveys of its DI callers in 2001. RECOMMENDATIONS To ensure that customers have access to the DI program’s telephone services, the department should take these actions: • Continue to explore and use the various methods allowed in the State’s civil service system to hire promptly and retain call center staff. These methods could include, but are not limited to, providing an ongoing pool of eligible candidates to hire from, reassessing the minimum qualifications, using part-time staff that can convert to full-time staff to fill vacated full-time positions, and offering training and devel- opment assignments to attract staff from other programs. 39 • Complete customer service contingency plans and work closely with its telecommunication contractors to limit the effect and number of system breakdowns during installation of future system changes. • Modify management reports to include the call activity of both its customer service units and its call centers and to measure and report its performance in meeting its goal. • Conduct its planned customer satisfaction surveys. We conducted this review under the authority vested in the California State Auditor by Section 8543 et seq. of the California Government Code and according to generally accepted government auditing standards. We limited our review to those areas specified in the audit scope section of this report. Respectfully submitted, ELAINE M. HOWLE State Auditor Date: July 10, 2001 Staff: Reed M. McDermott, CPA Russ Hayden, CGFM Matthew Liu Peter A. Foggiato, III 40 APPENDIX Summary of New Telephone Services’ Effects on the Unemployment Insurance and Disability Insurance Programs C hapter 329, Statutes of 1998, required the Bureau of State Audits to examine for several factors the effects that implementation of new telephone services had on the unemployment insurance (UI) and disability insurance (DI) programs administered by the Employment Development Department (department). The UI program provides to its customers claim filing and information over the telephone. The DI program provides information assistance only. We summarize the impact of new telephone services on each factor below. Effect on the Effect on the Factor Unemployment Insurance Program Disability Insurance Program Impact on customer service Customer services and access to program Although the department improved its information have improved. Nevertheless, one customer services and access to program of the department’s system modifications has information, many callers have encountered rendered the system unable to ascertain if the busy signals. Additionally, the telephone department is meeting its customer service system asks a significant number of customers goals. Several surveys of customers found that to call back later. Before the department they are generally satisfied with the services implemented its toll-free telephone number, they received. Chapter 1 discusses more fully callers stated in surveys that they were largely the impact on customer service for this satisfied with the level of services received program. over the telephone. The department plans to conduct other surveys of callers in 2001. Refer to Chapter 2 for further information about the effect of telephone services on this program. Customers’ ability to access After calling the toll-free telephone numbers, After calling the toll-free telephone number, routine and nonroutine customers can easily access routine and callers can easily access routine and information about their nonroutine information about individual cases nonroutine information about individual individual cases by using the interactive voice response system cases by using the voice response system or (voice response system) or by speaking to a by speaking to a customer service representa- customer service representative. However, tive. However, our analysis disclosed that for during peak periods of service, callers who wish the three months after the department to speak to a representative may encounter installed new equipment in January 2001, longer wait times than usual. Refer to Chapter 1 nearly 31,000 callers received a busy signal. for additional information. Additionally, during the 15-month period from January 2000 through March 2001, 27 percent of the callers who asked to speak to a service representative at the call centers were asked to call back later. Chapter 2 discusses the impact of this factor in more detail. 41 Effect on the Effect on the Factor Unemployment Insurance Program Disability Insurance Program Impact on the number of The department’s use of new telephone services The department’s use of new telephone appeals filed has had no noticeable impact on the number of services has had no noticeable impact on the claimant appeals filed. number of claimant appeals filed. Costs for telephone services When the department established toll-free When the department established the toll-free per number of telephone telephone numbers, it reduced the number of telephone number, it reduced the number providers telephone service providers. Thus, its actions of telephone service providers. Thus, its eliminated concerns about the costs of actions removed concerns about the costs of having numerous providers. We performed no having numerous providers. We performed no further review. further review. Ability of customers to The department requires written documents The department still requires each customer to provide written only for a few types of UI claims. Thus, most submit a DI claim form certified by their documentation to the callers need not submit any forms or written medical provider. Thus, the addition of Employment Development materials. The department ensures the accuracy new telephone services has had no impact on Department of customer records by sending in a timely DI program customers’ ability to provide manner the notifications to customers and the written documentation. most recent employer and by requesting that they make corrections to the records, if needed. When callers request copies of their records, the department sends the desired information within a specified number of days. Impact on customers with With its new telephone services, the With its new telephone services, the disabilities department eliminated many physical barriers department eliminated many physical barriers to service. Additionally, the department to service. Additionally, the department established a toll-free teletypewriter (TTY) established a TTY telephone number for its telephone number for its callers who have callers who have speech or hearing speech or hearing impairments. Callers who impairments. Callers who use TTY represent use TTY represent less than 1 percent of all less than 1 percent of all DI callers. UI callers. Effectiveness of language The department established toll-free telephone The toll-free telephone number that the services numbers for its callers who speak Spanish, department implemented includes a Spanish Cantonese, and Vietnamese. Spanish-speaking language option for its callers. The department customers can access information through the provides to its DI callers interpretive services in voice response system 24 hours a day and speak 34 other languages. to customer service representatives during regular business hours. The telephone system routes Cantonese and Vietnamese callers directly to representatives during regular business hours. The department gives its UI callers access to interpreters in 25 other languages. 42 Agency’s comments provided as text only. Health and Human Services Agency 1600 Ninth Street, Room 460 Sacramento, CA 95814 June 21, 2001 Ms. Elaine M. Howle State Auditor Bureau of State Audits 555 Capitol Mall, Suite 300 Sacramento, CA 95814 Dear Ms. Howle: Thank you for forwarding for my review and comment a draft copy of the Bureau of State Audits’ report entitled “Employment Development Department: Although New Telephone Services Have Enhanced Customer Access to Its Unemployment and Disability Insurance Programs, Customers Encounter Difficulties During Peak Calling Periods”. I am forwarding to you the EDD’s responses to the review findings and recommendations. I understand that the EDD has begun taking steps to address the issues raised in the Bureau’s report. Thank you once again for sharing the draft copy of your findings and recommendations. If you require further information concerning EDD’s call center system, please do not hesitate to contact me directly. You may also contact Michael Bernick, the Director of the Employment Development Department at (916) 654-8210 to assist you. Sincerely, (Signed by: Grantland Johnson) GRANTLAND JOHNSON Enclosure 43 Employment Development Department PO Box 826880 Sacramento, CA 94280-0001 June 20, 2001 Ms. Elaine M. Howle State Auditor Bureau of State Audits 555 Capitol Mall, Suite 300 Sacramento, CA 95814 Dear Ms. Howle: This letter responds to your draft report “Employment Development Department: Although New Telephone Services Have Enhanced Customer Access to Its Unemployment and Disability Insurance Programs, Customers Encounter Difficulties During Peak Calling Periods.” We feel that in general the review is balanced and appropriately identifies current and future challenges to our delivery of Unemployment Insurance (UI) and Disability Insurance (DI) services. We appreciate your recognition of the improvements we have made to give our customers better access. We agree that further improvements can be made in the use of telephone technology for service delivery and we feel that this technology is essential to our being able to provide services to our customers. For clarity, each Bureau of State Audits (BSA) finding is summarized below, with the corresponding recommendation statement(s), followed by the Employment Development Department’s (EDD’s) response. Unemployment Insurance Program Finding: Customers sometimes have lengthy waits during peak calling periods. Recommendation: Continue to explore ways within the State’s civil service system to hire part-time or seasonal staff to reduce customer wait times during expected busy periods. 44 Ms. Elaine M. Howle Page 2 Staffing has been of concern in the UI program, and we have used a variety of strategies to address this issue. With the approval of the State Personnel Board, the Department attempted to use the Staff Services Analyst (SSA) list as an appropriate list from which to hire staff. Inquiries to candidates on the SSA list proved to be less than successful, with only a small percentage of the candidates indicating an interest in our UI positions. We are now using a continuous filing process to ensure a continuous pool of eligible candidates from which to hire, and are currently reevaluating the minimum qualifications. Recommendation: Consider performing a study to examine the effect on call center workloads of increasing business hours for call centers during peak calling periods. The EDD will explore the feasibility of extending business hours during peak calling periods. The EDD’s automated system is only available from 6:00 a.m to 6:00 p.m. for claim filing purposes. During the remaining hours, the system is running batch processes such as UI checks and employer and claimant notices. The availability of the system allows us a three-hour extension of our current hours. Historically we have extended hours and days of service based upon the availability of staff. This is a recommendation we would use as we continue to hire more staff. In addition, several technical changes are planned to improve customer service. Our next planned improvement to the current telephone system will be the inclusion of a wait message. The wait message will advise callers of the anticipated wait time until a representative can answer their calls. While this message may not affect the number of abandoned calls, we believe that it will reduce the number of redials by customers who abandon and then immediately try to re-enter the system. The wait message will also reassure customers that a representative will, in fact, answer their calls. We have already established a workgroup to develop the anticipated wait time message. Finding: The Department cannot ensure that it has met the customer service goals it established for the program. 45 Ms. Elaine M. Howle Page 3 Recommendation: Promptly complete its process for setting challenging yet reasonable goals for answering customer calls. Additionally, it should modify its management reports to measure its performance in meeting these goals. The conversion to the Intelligent Call Routing (ICR) system allows customers to call a toll- free 800 number for both UI claim filing and information purposes. The ICR, implemented in 1999, provided the Department with information that identified access problems to the system: customers calling to file UI claims were being blocked by customers attempting to retrieve claim information. In response to this blockage, the Department installed additional lines and ports, providing increased access to the automated call distribution groups. These technical changes (ICR and additional lines and ports) resulted in better access for our customers, but were not fully operational until March 2000. With a complete year of data collection, through a full cycle of high and low workload periods in which the network system was sufficiently sized and essentially problem-free, the Department can now reasonably establish response time goals and modify management reports to measure performance. We plan to implement these changes during the 2001-02 state fiscal year. Additionally, in an effort to further expand access for our customers, we are developing an Internet application for UI claim filing. We expect to release this application into a testing environment sometime next year. An Internet application will provide access to our customers anywhere that they have access to a personal computer with an Internet connection. Recommendation: Conduct planned customer satisfaction surveys of teletypewriter users and Cantonese and Vietnamese speaking customers. The EDD plans to continue to conduct on-going customer service surveys to get customer feedback on UI services, including teletypewriter users and Cantonese and Vietnamese speaking customers. Disability Insurance Program Finding: Staffing shortages at call centers reduce customers’ access to Disability Insurance services. 46 Ms. Elaine M. Howle Page 4 Recommendation: Continue to explore and use the various methods allowed in the State’s civil service system to hire promptly and retain call center staff. Like the UI program, staffing has been of concern in the DI program and we have used the same strategies employed by the UI program. With the approval of the State Personnel Board, the Department attempted to use the Staff Services Analyst (SSA) list as an appropriate list from which to hire staff. Inquiries to candidates on the SSA list proved to be less than successful, with only a small percentage of the candidates indicating an interest DI positions. We are now using a continuous filing process to ensure a continuous pool of eligible candidates from which to hire. Additionally, we plan to transition workload from the five customer service units to the customer service centers, thereby expanding seating capacity and improving recruiting efforts to hire additional staff. Finding: The Disability Insurance program cannot fully monitor its performance in meeting its service level goals. Recommendation: Modify management reports to include the call activity of both its customer service units and its call centers and to measure and report its performance in meeting its goal. The DI program has identified, and the audit report confirms, the need to improve the management information systems in the five customer service units. As mentioned above, we plan to transition workload from the five customer service units to the customer service centers. This transition will provide consistent management reporting on all call activities. We plan to improve our call distribution and routing systems by exploring new software and telecommunication services to allocate calls more efficiently among our centers. Recommendation: Conduct its planned customer satisfaction surveys. The DI program plans to conduct on-going customer surveys to get feedback on customers’ satisfaction with DI services. Finding: System problems have prevented callers from receiving service. 47 Ms. Elaine M. Howle Page 5 Recommendation: Complete customer service contingency plans and work closely with its telecommunication contractors to limit the effect and number of system breakdown during installation of future system changes. DI Program management will actively partner with our telecommunication providers to develop and refine contingency plans, as well as thoroughly test system enhancements prior to implementation. We thank the BSA staff for their professionalism and openness. During the next few months we will begin to research and address each of the recommendations in more detail. As discussed in our exit interview of June 18, 2001, we will be providing follow-up reports on our progress in improving customer service levels. Our first report will be sent to you on approximately August 17, 2001. If you have questions about this response, please contact Pam Harris at 654-7014. Sincerely, (Signed by: Frank Collins for) MICHAEL S. BERNICK Director 48 cc: Members of the Legislature Office of the Lieutenant Governor Milton Marks Commission on California State Government Organization and Economy Department of Finance Attorney General State Controller State Treasurer Legislative Analyst Senate Office of Research California Research Bureau Capitol Press 49