Back to Department of Employment Services main page
|
Columns DCWatch
Archives Elections Government and People Budget issues Organizations |
FINDINGS OF AN INTERNAL INVESTIGATION REGARDING THE DISTRICT’S 2008 SUMMER YOUTH PROGRAMAugust 12, 2008 EXECUTIVE SUMMARY In a letter dated July 30, 2008, Mayor Adrian M Fenty made a request to the District’s Chief Financial Officer to use the District’s Contingency Cash Reserve Fund to increase the budget allocated to the city’s Summer Youth Program by $20.1 million, raising to $52.4 million the projected cost of the program. This increase came amid concerns that the program was grappling with a number of serious operational challenges. This report details these challenges that led to this increase, their causes, and strategies to resolving them in 2008 and learning from them in 2009. The major findings are:
Payroll for the fourth pay period will, for the first time since the program began, pay less than the maximum allowable hours. The approach is to: 1) identify and remove from the system people who are not part of the program at this point, and 2) ensure that payments reflect the program’s most fundamental parameters involving pay rates, age limits, and allowable hours. OCTO will continue to methodically gain control of payroll costs, and conduct a profile analysis to determine if a disproportionate number of problems concentrate at particular work sites. To learn from the experiences of 2008, this report makes the following recommendations: 1. Invest in programs that work, stop funding those that do not, based on a rigorous evaluation of program quality and outcomes. 2. Maintain a commitment to providing quality experiences to a large number of District youth. 3. Reassess the skills required to run the program, and put the right people in the right job. 4. Enforce time entry requirements at host organizations, and penalize those that fail to comply. 5. Have job assignment available for youth as they register, with goal of registering a higher percent of participants before the program begins, by securing contracts earlier or migrating to multi-year contracts with providers. 6. Demand that assumptions are relentlessly challenged, and train DOES staff how to respond to unexpectedly large workflows. 7. Eliminate all paper processes and bring the entire summer program online. 8. Continually validate information to stop fraud and abuse before or just after it occurs. 9. Integrate the debit card with the One Card to assist the city guard against future abuse. PURPOSE At the request of Mayor Adrian Fenty, this report conveys the findings of an internal investigation of the 2008 Summer Youth Program. Its objectives are to: 1) identify the problems associated with the program, 2) describe how and why they developed, and 3) describe actions underway to ensure they do not happen again. SCOPE AND METHODOLOGY To detail the full extent of problems related to this year’s Summer Youth Program (SYP), we included in our review events and decisions that date back to the 2007 SYP. We have also included pertinent information involving the planning process leading up to the 2008 program. This context is essential in understanding why problems emerged and why they persisted. The timeline and data included in this report continue up until August 11, 2008, or just after the end of the fourth pay period. Much of the information obtained for this report came from an examination of data extracts from the program’s registration system, time and attendance system, and payment system. We obtained additional information and insights through interviews and discussions with staff at the following agencies: · Department of Employment Services (DOES) · Executive Office of the Mayor (EOM) · Office of the City Administrator (OCA) · Office of Contracting and Procurement (OCP) · Office of the Chief Technology Officer (OCTO) · Office of the Chief Financial Officer (OCFO) We also conducted interviews with individuals at host organizations. It should be noted that the CapStat team has been reviewing the issues surrounding the SYP throughout the year. This report reflects that research, conducted over several months, as well as the work conducted since the Mayor’s request for a formal investigation was made. THE FLASH POINT In a letter dated July 30, 2008, Mayor Adrian Fenty made a request to the District’s Chief Financial Officer to use the District’s Contingency Cash Reserve Fund to increase the budget allocated to the city’s Summer Youth Program by $20.1 million, raising to $52.4 million the projected cost of the program. This amount represents an increase of more than 250% over the program’s original budget of $14.5 million. The urgency of this most recent request, made at risk of not being able to meet payroll for a workforce of as many as 21,000 young people, understandably raised a number of serious questions about the program’s management and operations. It also brought to the surface allegations of waste, fraud, and abuse. How could this happen? How did it happen? And, how will District government resolve the causes of these problems? The answers required our research to look back more than a year. THE PROGRAM AND THE ORGANIZATION SUPPORTING IT Each year for nearly 30 years, District government has administered a summer jobs program for teenagers and young adults, ages 14 – 21, who live in the District. The size of the program has fluctuated over the years, but its purpose remains the same: provide jobs and enrichment experiences for young people who may not otherwise find employment or training during the summer.
sAt one point, the program employed more than 20,000 young adults. But by 2003, it reached a low point in size, employing about 5,500 youth for a six-week period. Beginning in 2004, the program incrementally, though not consistently, expanded until 2007, when the city provided jobs, training, and enrichment experiences for more than 12,629 youth. Even with this near-doubling in size, there was widespread belief among those interviewed that this did not go far enough in offering opportunities to out-of-work, out-of-school teenagers and young adults. The Department of Employment Services’s Office of Youth Programs manages the Summer Youth Program, which is also called the Passport-to-Work Program. This office, composed of more than 40 FTEs including four managers, is responsible for planning and implementing the program each year. They also administer a much smaller year-round job training and employment program. PROGRAM GOALS FOR 2008 Formal preparations for the 2008 SYP program began in November 2007, and planning for it began even earlier. The objectives driving these efforts had their roots in lessons learned from the 2007 program. The 2007 program ran for eight weeks (it was originally scheduled to run for six, but was extended mid-way through the program), and begin on June 26. The new leadership at DOES, was able to incorporate several reforms in time for 2007. Most notably they made a push to increase enrollment after a year of decline in 2006, extend the program’s length beyond its original end date, and expand the variety of summer experiences available to youth. As early as May 2007, before the program began, DOES made a list of reforms that they had time to implement for 2007, and others that they planned to implement in 2008. This list focused largely on increasing program quality and efficiency, including more job- and life-skills training, online registration, and a call for greater involvement from the private sector. Later that year, between July 2007 and the end of the fiscal year, DOES crafted a FY2008 performance plan that set several additional objectives for 2008. Specifically, their performance plan called for DOES to “increase the size, duration, and quality of the Summer Youth Employment Program.” The plan set specific benchmarks of expansion: from 12,629 participants in 2007 to 15,000 participants in 2008 and from a 6-week program to a 10-week program. To lessen funding pressures, the agency would attempt to secure more private sector funding. Other ideas were developed and incorporated into the agency’s planning process well into 2008. Of all the proposed program improvements for 2008, there were four that proved critical to the story of how and why events unfolded as they did. The goals, some proposed by DOES and others by the Mayor and City Administrator, were to: 1. Dramatically increase enrollment, so that no young person needing a summer job or experience would be turned away, no matter when they registered; 2. Expand the variety of opportunities, and thus the number of work sites managed by the program; 3. Change the payment method from an electronic benefit transfer (EBT) system to a debit card system, which would allow young people in the summer program to have quicker access to their money and not be forced to pay ATM and other bank fees, as they had with the prior system; 4. Implement electronic time and attendance, intended to improve the accuracy of payments and create organizational efficiencies by eliminating paper records and manual entry. FINANCES AND BUDGETING Effective financial planning for the Summer Youth Program requires accurately forecasting five variables: 1) the number of young people who register for work; 2) the ages of these participants (since the maximum hours one is permitted to work per week varies by age); 3) the hourly pay rate; 4) the length of the program; and 4) attrition, or the rate at which youth leave the program voluntarily prior to its official conclusion. The approved FY2008 budget contained $24.5 million in funding for Youth Programs at DOES. This sum includes funding for both DOES’s summer employment, indirect administrative costs, as well as several other youth-oriented programs, including the agency’s year-round program as well as agency indirect/overhead costs for program management. Of this, approximately $14.5 million was proposed and approved for SYP salaries and contracting costs. The sequence and timing of the budget cycle is essential in explaining subsequent events. The Mayor submitted his FY2008 budget request on March 23, 2007, prior to when DOES committed to many major policy goals listed previously. The proposed and approved budget assumed a 6-week program employing about 11,000 young people. On May 15, 2007, Council approved the city’s FY2008 budget. Ten days later, on May 25, President Bush signed into law legislation increasing the federal minimum wage, which is the basis for the program’s pay rate. Later in the year, DOES increased their employment goals to 15,000 young people. The financial impact of these two decisions led to a supplemental funding request approved by Council on December 18, 2007, that increased by $7 million the approved budget of DOES’s Youth Program, and increased by $14.5 million the agency’s overall budget. By the time the program began in June 2008, funding attributed to the SYP increased from $14.5 million to $21 million. Nevertheless, the increases continued, as the program’s budget was sufficient for only 15,000 participants. The third of four budget increases occurred when the Mayor requested a reprogramming, which Council approved on July 17, 2008, in the amount of $10.8 million, bringing the SYP program funding to $32.3 million. The administration attributed this increase to the unanticipated effectiveness of recruitment and outreach, which resulted in more than 21,000 registrants, higher than the original or revised estimates for the program. The most recent and largest budget increase was requested just two weeks later, in the amount of $20.1 million. This request, which would bring the budget up to $52 million, is equivalent to the amount required to pay for all the young people who signed up for the program as if there was perfect attendance. In other words, this assumed that all the young people who were assigned to a worksite would be paid for working the maximum allowable hours for the entire duration of the program. The failures that led to the use of this assumption in payment and budgeting is the central focus of this report. REGISTRATION The effectiveness of outreach efforts to register young people for the SYP was unprecedented in the recent history of the program. In 2003, the program employed 5,500; by 2007, enrollment exceeded 12,000. At a CapStat session prior to the 2007 SYP, the Mayor gave direction to DOES to improve their marketing efforts to reach youth who would not normally enroll in the program. Formal recruiting began as early as November 2007, when staff from DOES engaged in outreach and interviewing of college students as they returned to DC for holiday breaks. It continued throughout the winter and spring, as staff made efforts to attend community meetings and find other means of outreach and recruitment. The effectiveness of outreach efforts to register young people for the SYP was unprecedented in the recent history of the program. In 2003, the program employed 5,500; by 2007, enrollment exceeded 12,000. At a CapStat session prior to the 2007 SYP, the Mayor gave direction to DOES to improve their marketing efforts to reach youth who would not normally enroll in the program. Formal recruiting began as early as November 2007, when staff from DOES engaged in outreach and interviewing of college students as they returned to DC for holiday breaks. It continued throughout the winter and spring, as staff made efforts to attend community meetings and find other means of outreach and recruitment.
Predicting the final number of youth enrolled became a challenge, since the impact of the intensified outreach efforts was not entirely known and there was no cut-off date for registration. Staff expected a large number of last-minute enrollees, but did not foresee how large that group would be. In March and April, DOES believed that demand among youth people for summer employment would exceed even the revised targets of 15,000. Because of an aggressive spring recruitment campaign, DOES estimated that a large number of youth in need of employment would likely not turn their attention to seeking it out until the school year wound down. To accommodate this interest, the Mayor gave direction for DOES to allow any young person who could demonstrate DC residence to enroll in the program, without artificial registration deadlines. As late as early June, DOES senior staff believed that the program’s final registration would plateau at between 16,000 and 17,000 youth, but they could not be sure how significant a last-minute surge in registration would be. Finally, in the second week of June, the decision was made to include into the Summer Youth Program summer school students at DCPS who were engaged in educational enrichment. Youth were not paid to attend school. Rather, they were placed under the supervision of a program coordinator after the completion of their classroom work each day, and given work and enrichment assignments that complemented their academic requirements. The rationale behind this decision was to eliminate the incentive for a young person to leave summer school in order to take a paid position in the SYP. By August, DCPS will have doubled (to 443) the number of students who graduated from their grade during the summer. The integration of these programs increased enrollment by an additional 2,300 students. Both the timing and size of program registration combined to overwhelm program administration and was a significant factor that led to disarray later in the program. The dramatic expansion in the jobs program – nearly doubling it in size in a single year – proved beyond the organization’s capacity to accommodate and also execute the program’s core operations. CONTRACTING AND NOTIFICATION On March 3, 2008, DOES put out a solicitation for project proposals from potential summer youth host organizations. At the time, the solicitation closing date was April 3, which would have provided ample time to enter into contracts with the winning responses. In fact, DOES maintained detailed contracting target deadlines when it began process. DOES and OCP managers met prior to the solicitation to review language for compliance with contracting regulations. According to interviews with an OCP staff member involved in the project, the city received more than 30 proposals in response to the solicitation. The OCP project manager at the time believed that more time was needed to answer questions raised by prospective hosts. He then extended the closing date to April 23. Less than a week after the closing date, the OCP manager assigned to the contracts retired from government service. Next came the negotiations period. During this time, contracting officials called organizations whose proposals were in the “competitive range” to discuss possible proposal edits and revisions. OCP then told the organizations that their “best and final” proposals were due May 8. Three days prior to this date, a second OCP employee assigned to the project left, in this case for medical leave. At the time, she believed that the schedule was consistent with prior years and on target for completion. DOES evaluated the submitted proposals and on May 21 sent a report on the proposal evaluations to OCP. The following day the contracting office sent back questions about the report to DOES, and evaluations were re-conducted May 23 – 26. Contracts were drafted on May 29. Just prior to finalizing contracts, however, the OCP manager who was now assigned to the project believed that errors and inconsistencies existed in the original solicitation, despite the fact that it had been previously approved. Since putting out another request would take too much time, OCP decided to amend the existing solicitation and on May 30, with just over two weeks left before the program began, discussions were reopened to all organizations with proposals in the “competitive range.” This round of discussions closed on June 2, the “best and final” proposals from these negotiations were due June 3, and the evaluation of these proposals was conducted on June 4. By contrast, the RFP closure date in 2007 was April 20. On June 5, OCP sent contracts to the organizations with the highest-scored proposals for their signatures. The number of contracts sent out on June 5 was based on the amount of money budgeted for the summer youth host organizations at that time. On June 6 DOES told OCP that additional funding had been identified and on June 9 contracts were sent out to additional organizations. OCP received the first signed contract on June 10, and all 36 contracts had been signed by June 16. The late completion of contracts was a critical event. There were 36 contracts identified during research, totaling more than $10 million and enrolling approximately 6,800 youth, which were finalized between June 10 and June 16. In fact, 8 contracts were awarded on June 16, the first day of the program. Furthermore, these awards were all made on a contingency basis, meaning that actual Purchase Orders were not established until a few days later. Several purchase orders, consequently, were not approved until after the program started. From DOES’s standpoint, despite a plea by its senior executives, the process did not move quickly enough. What makes the contracting process so significant is that DOES’s ability to assign registrants to worksites is fundamentally dependent upon it. DOES could only assign youth to work sites when valid contracts existed with the work site host. If a large portion of its contracted capacity was tied up in the procurement process, then a large number of registered youth would remain unassigned to job sites.
One week prior to the start of the summer program, approximately 12,000 registered youth had not been told where to report to work. DOES had prepared form letters to be mailed out during that week to notify these youth of their assigned location, as they had been doing previously. As contracts were awarded, kids would be notified of their placement. For reasons still not known, only about 7,000 letters were actually sent out. Although the Director had been told that “letters had gone out,” she was not told how many, or more importantly how many were not, sent out. Even if all letters had been sent, there would still be several thousand youth who would be assigned so late in the process that mailing was not a realistic option. In a last attempt to reach all the youth still unaware of their assignment, DOES placed a “robo-call” to inform the young persons and their families of a website that DOES established where they could find out their work assignment. Even if this attempt had proven successful, disarray would have emerged at the start of the program since there were still several host organizations with unapproved contracts entering the weekend and there is a clear differential in access to the internet throughout the city. In short, as the program began, youth did not know where to go to work, and DOES did not know where to send them. THE TIME AND ATTENDANCE SYSTEM Prior to 2008, DOES maintained a general purpose time and attendance IT system called the Virtual One Stop. The system worked as follows. During the pay period, site managers (those supervising youth at their jobs) would record time and attendance on paper. At the end of each pay period all sites would submit these paper timesheets to DOES’s Office of Youth Program payroll office. Worksites sent timesheets to DOES in one of two ways. Hosts either faxed them into DOES from the site, or the worksite would send the original timesheets via a courier service arranged by DOES. The “payroll office” within the Summer Youth Program consisted of one full-time DOES employee who managed a team of summer youth, to whom DOES gave the responsibility for entering time for the thousands of fellow summer youth employees as the basis of their pay. A manual, remotely collected time entry system had obvious flaws. It required manual-entry timesheets, timesheets of varying clarity, required a manual transfer of the document via fax and courier, and then relied on summer youth to enter time accurately for a workforce that was roughly one-third the size of the entire DC government. There were ample possibilities for data errors and inadequate auditing controls. To alleviate these risks, DOES’s IT Office set out to replace it with a web-based time and attendance system that would allow supervisors to enter data electronically at their worksite. The system would then compile this information in a way that it could be sent directly to financial institutions as the basis for payment. What began as a sound policy decision fell victim to poorly managed execution. DOES designed and began development of a new time system without sufficiently informing or seeking guidance from OCTO, the city’s central IT office. This lack of communication, in hindsight, cut off expertise and resources that would have ensured the quality of the program and sped up its completion. The DOES IT office completed the new application just 2 weeks prior to the start of the SYP. With so little time for implementation, testing for errors and other bugs was minimal. Minimal to no testing was provided by OCTO, since they were not involved in the system’s development until after the program began. Of equal importance, there was almost no time for training and genuine buy-in among both DOES’s staff and host organizations. With less than two weeks left before the SYP began, the DOES Director discussed options on how to migrate over to the new system with her senior staff and the Office of Youth Programs. The Director was concerned about the mounting work required to accommodate the increasing flow of new registrants and host organizations into the program. The Director pushed off system implementation for another few days, until after the weekend of June 7 and 8, so that staff could continue enrolling youth into the old system, with which they were familiar and therefore could more efficiently use. On Monday, June 9 and Tuesday, June 10, a week prior to the beginning of summer, DOES’s Youth Programs Staff were trained on how to use the new time and attendance application. Training the staff at host organizations and work sites to use the new system proved more problematic on such short notice. DOES rolled out daily webinars to train host organizations on the new system. This was also problematic. The likelihood that these webinars would actually reach everyone who needed to learn the system was small. Second, an undermined but sizeable number of sites were not equipped with the technology to accommodate web-based time entry, and the short notice left virtually no time for upgrades. Knowing this, DOES maintained their paper-based system, which would run identically to the old method, allowing work sites to send in paper attendance forms to be entered into the new time and attendance application at DOES’s central office. Inadequate testing and data verification in this new time entry system proved to be even more damaging than the lack of training. Based on interviews with DOES managers, it appears that a significant amount of outdated information was transferred from the Virtual One Stop system to the new time and attendance application. For example, some youth registered but were not assigned to a work site. Youth and work sites from prior years would also show up as active in 2008. In tests run by OCTO after the program went into crisis, however, they concluded the system overall was sound. It is still unclear if reported problems tied back to system design or how the staff used it. Regardless of this, there was no time to diagnose these issues and make adjustments prior to the beginning of the program. THURSDAY JUNE 12: PERCEPTION VERSUS REALITY On Thursday June 12, 2008, the Mayor convened a CapStat session to receive a final briefing on the SYP before the program’s start. CapStat is an accountability program for reviewing agency performance and providing the Mayor and City Administrator with accurate data about how effectively programs are implementing major policies and carrying out their core operations. In one-hour accountability sessions, the Mayor and City Administrator bring into one room all executives responsible for improving performance on a single issue, examine performance data around that issue, and explore ways to improve government services, as well as make commitments for follow-up actions. The session scheduled for June 12, entitled Summer Readiness, looked at the status of preparations at four different agencies: the Department of Parks and Recreation, DOES, DCPS, and OCTO. In preparation for sessions, the CapStat team provides all participants with statistics and graphs that highlight performance trends. The charts for DOES focused on registration, funding, and placement. At no point in the session did the severity of the problems emerge. The DOES portion of the discussion went smoothly. When the DOES director said job sites would enter students’ time online, for instance, no questions were asked about job sites’ capacity to do this or what would happen if the online system failed. The Director did mention that DOES had plenty of time – a weekend, Monday, and part of Tuesday – to correct any payroll errors after the first work week, but did not discuss a plan for mobilizing employees if any major payroll glitch occurred. Job placements were not discussed at all. At one point the City Administrator asked about capacity versus enrollment, but it was treated as an aside and the discussion quickly turned to the OneCard. In part because of the major increase in enrollment, what DOES conveyed to the Mayor and City Administrator was a program on the verge of success. More scrutiny should have been given to DOES by the CapStat team to verify information and provide enough skeptical pushback to ensure that the Mayor and City Administrator received an honest appraisal. In retrospect, DOES built a bigger program on an existing foundation – last year’s successful program and expansion –which was subsequently found to have reached its capacity for growth without significant organizational change. By contrast, the CapStat team and city executives provided far more scrutiny to the Department of Parks and Recreation for the same session. The duration of questioning directed at the DPR Director was nearly twice as long as that given to DOES. The CapStat team requested direct data extracts of registration figures for programs, conducted mystery shopping to see if waiting lists existed for programs, and even sent a representative out unannounced to outdoor pools to verify the number of life guards present and their names. Consequently, the toughest discussion of the session centered not on DOES, but on DPR. Apparently, DOES leadership had not pursued a deeper analysis of program structure, performance, and readiness beyond that offered at the CapStat. This management presumption, of trusting but not verifying, proved inadequate. The DOES director reported meeting with her SYP staff every day for 30 days prior to the program’s start, but took at face value information that her staff presented. There was even a detailed project plan, replete with deadlines, milestones, and project phases, loaded into Microsoft Project that planned nearly every detail of the operation. Most of the basic facts surrounding the planning process, which have been described in this report, were known by the Director. To roll out the program without crisis, given the information available at the time, would have required DOES to implement every aspect of its operation flawlessly, demonstrate extraordinary adaptability, and have many external variables fall into place. As major problems overwhelmed the program, DOES did not seek the help it needed. This collective trust in the best-case scenario was in stark contrast to the events that were about to unfold. FRIDAY JUNE 13: CHAOS AWAITS In reality, disarray was unavoidable by Friday, June 13, the day before the beginning of orientation and the last business day before the program began. The individual planning processes described previously collided in a way that compounded the damage of each by itself. DOES had 17,244 registrants on record, but by mid-July would have 21,073. DOES had planned for 14,341 placements and budgeted for 15,000, reflecting an initial attrition rate of 15%. In fact, the program would place more than19,000, an initial attrition rate of only 9%. The rush of interest among young people looking for opportunities then needed to be connected to a relatively untested time and attendance system that was entirely new to its users. As planned, every young person was supposed to know their assigned job placement and location of work before summer began, but due to late contracts and last-minute enrollments, there were an unknown number of youth – though certainly in the thousands – who on June 13 had no idea where they should report for duty the following Monday. Over the next few days, thousands of young people would converge on DOES’s headquarters on H Street, NE seeking assistance. JUNE 14 – JUNE 20: THE FIRST WEEK On Saturday June 14, the DOES Director went to her H Street offices to pick up materials to bring to the Convention Center where orientation was about to take place. She was greeted with the sight of perhaps as many as 1,000 youth lined up outside DOES’s doors. The young people that came to DOES that Saturday, then continued to return for the next several days, consisted of those who were interested in registering for a job, others who had registered previously but did not know their site assignment, and a few whose assigned work assignment was nonexistent. An atmosphere of chaos, tension, and eventually exhaustion persisted throughout the first several days, according to those who worked the site. Another theme that emerged during interviews was a lack of clarity about how to handle the large numbers. Planned workflows were disregarded, in many cases because they were neither designed for nor capable of handling the demands of those first few days. The operation dealt with the problems before it, with little focus on cause. One worker at H Street captured her observations: On Monday, June 16, I became involved in directly working with the youth. I would receive youth who were: registering, getting work assignments, and transferring assignments. There was a line around the building, people had been coming in through the weekend. There were kids moving all around the first floor. Once I received an account I was able to do all the activities. There were several issues with the IT system, things as basic as scrolling through work sites. In some instances I was unable to de-couple a youth from a site to transfer to another site. I would usually persevere and get it done, which usually required unlocking by IT; however I could hear people saying they couldn’t do it, so they would handwrite a job site for the youth and send them out.
Managers were also worried that they might not have enough worksites to handle the unprecedented demand for jobs. Just as much a focus of the team was finding or confirming additional placements at worksites. The vast majority of these new placements were filled successfully. But, workers did report youth who were turned away from a site because of overcrowding, were sent to closed sites, and in a few cases returned multiple times until they found a job. Some employers adapted to an unexpected increase in their youth workforce better than others. In some cases, particularly placements inside District government, the increase was dramatic. Logistical challenges around finding space and equipment were common for these agencies, as was the task of creating additional projects for their summer youth. No program-wide analysis has been conducted on the extent to which the quality of the experience was diluted as a result. Anecdotally, though, agency responses varied. It should be noted that some host organizations felt that the sudden expansion was, in fact, not a challenge; they had projects that could increase in size easily. For others, there were difficulties identifying quality work assignments. Each evening, beginning on Saturday and continuing virtually nonstop for the subsequent weeks, a team of mostly DOES staff would work well into the night and early morning trying to collect and process the information and forms collected during the day. Among the registration forms were some that were dated back to March, April, and May, but for some reason were never entered into the system, or were initially collected by a third party and not brought to DOES until mid-June. In other cases, DOES had young people who had registered, had proof of registration, but were not in the system and were not referred to a job site. It is unclear if these observations reflected a broken IT system or an organization that had failed to enter the registrations that they were given. On Monday, June 16, and each day throughout that work week, hundreds of young people continued to come to H Street for registration and to the Convention Center for orientation. At orientation, each young person would receive basic workplace training, financial training, receive a OneCard for identification, and a debit card for payment. Training occurred in four-hour segments of several hundred youth per segment to cycle through all registrants. Only the debit cards for those kids scheduled to attend a particular orientation were available for dispersal. At the time, this policy was intended to maintain controls over the deployment of debit cards. But amid the confusion, many youth went to the wrong orientation session, or none at all. These youth started their job without a payment mechanism in hand. At this point, staff from the Executive Office of the Mayor and Office of the City Administrator became directly involved in operations, trying to alleviate the workload, whether to provide management support, data analysis, or simply provide additional bodies to staff intake stations. The involvement of the OCA and EOM remains with the program to this day, and in the past two weeks has assumed an increasing degree of control. The primary focus of those working at H Street was to move as many kids as possible through the building. Amid the disarray and need to work efficiently, standard procedures were replaced by ad hoc ones, which might have varied from person to person. Electronic data capture was replaced with handwritten notes, not always recorded on proper forms, but in some cases written on any sheet of paper that was available. There is little evidence to suggest that these lists were ever compiled and less to suggest that any analysis of trends, root causes, problem sites, or recurring issues were identified from this data. Basic standards and controls to guard against abuse broke down. Several workers – some DOES and others pulled from other organizations – recorded and handled all the information needed to register, including addresses and social security numbers. This information, in many cases written on loose pieces of paper, was needed to process the youth. Again, the pressure to cycle through a large numbers in a short timeframe overwhelmed many concerns that would normally halt an operation. One worker described the work process from her experience:
By every measure, however, the most damaging consequence of the chaos surrounding those first few days was that in the rush to assign kids to worksites, no one captured accurately all the work assignments that were made. In short, DOES did not know who worked where all of its summer workers were. Without this information, the city could not accurately capture time and attendance and thus could not accurately pay its summer youth workers. CHASING PAYDAY: THE FIRST PAY PERIOD The end of the first week of work also represented the end of the first pay period. It was clear to all the senior staff involved that after the events of the first week, processing complete and accurate time and attendance data would be impossible. Work sites could only enter time for youth they were officially assigned. With so many unrecorded assignments and so many worksites unprepared to use the new time entry system, it was not possible to have a high degree of confidence in the data that was entered. This left the administration with an unwelcome choice: either it could pay according to official time and attendance records, which would have inevitably underpaid or not paid thousands of youth who legitimately worked during that first week, or it could pay everyone registered in the system as if they had worked the maximum hours, which would have certainly paid some youth for more hours than they actually worked, since it is unlikely that every youth went to work every day. This latter option would assume perfect attendance, account for different age-based weekly hour limits, and given the circumstances was seen as the only option that would ensure that all those who did work received a paycheck. The administration chose to pay all the kids. In explaining the rationale behind this decision, one interviewee summarized it this way: “We did not want to punish kids for mistakes made by adults.” Even this, according to interviews, did not prevent complaints of underpayment on pay day, in part because DOES had not yet distributed a large number of debit cards (there were 11,966 cards activated, and an additional 9,283 cards issued but not distributed1). DOES set up a pay-day processing center at the Convention Center to handle claims of incorrect payment. Working with the OCFO, the city had on-site about 9,300 undistributed debit cards, as well as a stack of checks with amounts filled in but blank name lines, along with CFO staff certified to process payments (164 checks were actually used). Youth who had not received a debit card would leave with one in hand. Those who lost their debit card would have their original one cleared of any balance, have another reordered, and because this process took 72 hours, would leave with a check in hand. THE SECOND PAY PERIOD If mistakes were corrected at that point, then the budget would not have ballooned to the extent it did, because historically attrition does not become a significant factor until later in the summer. The second pay period, however, led to the same result as the first. In contrast with the first pay period, however, the initial intention was to pay according to data reflected in the time and attendance system. But many registered participants and many records were still missing from the payroll system. DOES also did not set up a processing center after the second pay period, believing the data to be relatively clean. When pay day hit, kids showed up at DOES’s H Street headquarters with claims of underpayment or non-payment, claims that DOES considered likely to be true. There was, however, no way to verify each and every claim since DOES still did not trust data in the time entry system. Left with the same choice as before – to err on the side of underpayment or overpayment – the administration chose once again to pay all kids for the maximum allowable hours. Why did data integrity issues persist weeks after they occurred? Why did the payroll problems continue? Based on discussions with managers and other staff, two reasons emerged: 1) everyone involved, but most notably DOES, consistently underestimated the extent and complexity of the problem, and 2) DOES did not demonstrate the organizational competence to methodically plan and execute an appropriate response to these challenges. As a result, too few people, and in many cases people with the wrong skill sets, were deployed to resolve the crisis. Help was requested and given, but not enough help, and more importantly, not the right kind of help. Everyone’s energy was on reacting to the crisis immediately in front of them, rather than solving the core problems. In conducting research, several examples were presented that demonstrated this. One person described how ideas were proposed about calling all worksites to verify names in the time entry application, but were never followed through. Another interviewee described how worksites were allowed to submit time and attendance forms up until only a few hours before data needed to be sent to financial institutions. This left too little time for identifying and correcting problems. In yet another example of the workflow challenges, an interviewee described the following scene at a payment center:
Allegations of poor work execution also consistently emerged in interviews. Some allege that work Youth Programs staff did not enter time entry forms that had been submitted. Others allege that clear directions were not followed by several staff members. Research was unable to verify whether and the extent to which willful insubordination occurred, but it should be noted that it was raised by several people. Others, who did not agree with allegations of widespread noncompliance, describe a workforce that by the second and third pay period was exhausted, and not up to fixing the crisis without additional help or replacements. THE THIRD PAY PERIOD: ASSESSING THE DAMAGE Shortly before the end of the third pay period, the City Administrator convened a meeting with nearly all the key managers to create a plan to gauge the depth of problems with payroll. Gathered around the table was staff from DOES, OCA, EOM, OCTO, and the OCFO. Enough data was obtained from the first two pay periods to indicate that the SYP still could not account for the placements of all its workers. The City Administrator laid out a plan that would call on a manual field operation to obtain attendance information from work sites that did not report any information via the time and attendance system during pay period two. This approach would serve as a gap analysis, to determine the difference between identifiable hours, staffing, site quality found through field visits, compared with payroll data collected electronically during the first two pay periods. The project managers assigned to this task pulled staff resources from throughout EOM, the OCA, and city agencies to deploy to the field. In all, roughly 60 DC employees went to the field over a three-day period to collect paper time records, and bring them back to a central processing center. The processing center, which was now moved to the Wilson Building, was staffed 16 hours per day to enter time as it returned from the field.
These efforts were not designed to gain control of payroll, but rather help with that task later on. An analysis of data showed enough entry errors and inconsistencies to not gain the government’s trust. Once again the data complexity of challenges around time and attendance proved too formidable. For the third consecutive pay period, the city decided to pay all the kids the maximum allowable hours. THE FOURTH PAY PERIOD: GAINING CONTROL The fourth pay period ended on August 1. This report has been written between the end of the fourth pay period and the pay day for that pay period. It was during the fourth pay period that the city finally began to gain control of attendance and payroll information. Several other changes allowed for this. First, management of the crisis became a team effort involving several parts of government. DOES’s IT office was brought under managerial control of OCTO, the city’s technology office, and DOES’s IT Director was replaced. A new project manager was assigned by OCTO to manage all aspects of the crisis, and a “war room” was established at OCTO headquarters, and staffed by data mining and project management experts. Additionally, EOM began to pull appropriate staff from throughout District government to supplement, and in some cases take over, core functions of DOES’s Youth Programs Office. The strategy to gain control has been one of incrementally verifying and correcting entire classes of data error. More specifically, the team developed a comprehensive listing of every way that payroll information could be incorrect, then validated the likelihood of error based on testing against available data. This has allowed them to identify what errors are likely to exist, the financial impact of those errors, and then focus their attention on those that are of the highest risk. As each type of error is accounted for and resolved, the amount paid to youth can begin to reflect the real hours, pay rates, and participants. This process, as it continues to progress, has been critical in verifying the existence and extent of waste, fraud, and abuse. SHORT TERM ACTIONS (1): REDUCE PAYROLL Payroll for the fourth pay period will, for the first time since the program began, pay less than the maximum allowable hours. The approach taken to start fixing payroll for the program’s remaining weeks is to: 1) identify and remove from the system people who are not part of the program at this point, and 2) ensure that payments reflect the program’s most fundamental parameters involving pay rates, age limits, and allowable hours. More specifically, the focus is to: 1. Tighten our level of confidence in payroll data by: a. Validating all major sources of error, the findings of which were highlight in this report, and stop payments to individuals whom we have a high degree of confidence are no longer in the program b. Implementing sanity checks on payrolls prior to look for known issues that cause re-work at payday processing centers c. Eliminating the practice of maintaining information off-line, and bring online all prior records 2. Implement strong external communications with participants and their parents by: a. Contacting all individuals impacted by changes to payment strategy through multiple communications channels such as robo-calls, text messages, use of an outbound call center, and others b. Establishing a call center that will allow for many payment disagreements to be resolved without requiring that a participant travel to a processing center 3. Strengthen Pay Center readiness by: a. Opening pay centers in advance, which will result in problem resolved in advance and shorter lines This plan will scale back from the perfect attendance presumption in the following ways:
SHORT TERM ACTIONS (2): IDENTIFY AND RESPOND TO WASTE, FRAUD, AND ABUSE The full extent of waste, fraud, and abuse is not yet known, but significant progress has been made in gauging the likelihood and extent of it in the program. In some cases, allegations of fraud and abuse have not borne out. But in many cases, some presence of fraud has been verified. Collectively, there is great concern that hundreds – and perhaps thousands – of individuals registered in the youth program have been knowingly engaged in fraud. To be thorough, the report examines each known allegation separately. They are as follows:
SHORT TERM ACTIONS (3): THE CONTINGENCY RESERVE By the time of the third payroll, the program had exhausted its revised budget. Under the local Anti-Deficiency Act only two options existed: either an external source of funding had to be identified and made immediately available to cover cost overruns; or the program had to terminate immediately after only five weeks, ending summer employment for thousands of District youth who had registered, worked, and planned to continue to do so for the remainder of the summer. Cost estimates for a worst-case scenario (in which there is no improvement in the program's operational problems, most importantly the ability to verify time and attendance and thus deduct attrition from payroll expenses) projected that the program could exceed its revised budget by $20.1 million. On July 30, the Mayor formally requested an allocation up to this amount from the Contingency Reserve Fund, while efforts to reduce the cost of the program continued. The Contingency Reserve is a critically important component of the District's General Fund, created to meet unforeseen needs that arise during the fiscal year, including revenue shortfalls, unexpected obligations caused by changes in Federal law, natural disasters, and other unanticipated events with budget implications. As of December 31, 2007, the date of the most recent official certification, the Contingency Reserve fund balance was $223.5 million, including a receivable of $53.1 million. Any allocation from the Contingency Reserve must be repaid within two years. As noted in the Mayor's July 30 letter, it may be possible to repay the Contingency Reserve from the existing, certified fund balances of Special Purpose (O-type) accounts managed by DOES. Two of these accounts (0611 and 0624) were created to provide for the administrative costs of managing other employment-related programs. Because revenue collections have greatly outpaced recent spending, the balances of these administrative accounts nearly tripled from $7.7 million in FY 2004 to $21.4 million at the end of FY 2007. In addition, the balance of an account (0612) created to hold UI interest and penalties grew fivefold, from just under $208,000 in 2004 to nearly $1.1 million in FY 2007. A fourth account (0614/0600), with a balance of $1.4 million, has been dormant (with no revenues or expenditures) since FY 2005. While the balances in these accounts appear to be sufficient to repay the Contingency Reserve allocation to DOES, doing so would require a legislative change to the respective statutes that created the accounts. Given the administrative missteps that have caused the cost overruns in the summer youth program, it may be appropriate to repay the Contingency Reserve from these DOES administrative accounts. WHAT WORKED Program successes are not this report’s focus, but also should not be lost or forgotten. In fact, they need to be recognized as a fundamental to next year’s planning. In the course of researching problems, several innovations and improvements show demonstrable progress. Among those are: o A significant increase in the number of hard-to-reach youth were provided a summer job than in prior years. This factor contributed to the managerial challenges, but in itself should not be dismissed as without value. o DCPS more than doubled the number of summer school students who graduated due to their summer studies, which they attribute largely to not losing summer students to non-academic jobs o Juvenile arrests for violent crimes 2008 are down 34% (148 in 2007, 98 in 2008), while arrests of adults for violent crimes are up 4% (599 in 2007, 620 in 2008). Juveniles arrested for stolen vehicles are down 60% from 2007 (73 in 2007, 29 in 2008), while adult arrests for that crime are down 20%. o Introduction of the debit card system that allows participants to avoid bank fees and gain quicker access to their money. o Introduction of the One Card, a form of identification for program participants that also contained a Metro SmarTrip chip, loaded with $10, to make commuting to jobs easier. o Success of several large programs with defined objectives and enrollment, such as the District Department of the Environment’s Green Team. This program consistently retained a workforce of 250 individuals over all pay periods to date, who cleaned up neighborhoods surrounding rec centers, while receiving job training in emerging “green economy” industries, such as green roofs, tree maintenance, and water quality monitoring. o A second effective program was run by the Department of Youth Rehabilitative Services, which enrolled 260 youth. Youth enrolled in gained knowledge and skills that they can reflect on their resume as well as apply to future life experiences. Youth participated in landscaping (on grounds and within the community), where youth planted vegetables and mowed lawns for elderly residents; arts craft corps, where youth made murals and signs for the new facility; media advocacy, where youth learned the art of photography; outdoor leadership, where youth build boats and rope courses; and culinary, where youth learned the art of cooking and food preparation. Such activities not only gave youth an opportunity to give back to the community, but enhanced their leadership, conflict resolution, and communication skills. LONG TERM ACTIONS: RECOMMENDATIONS FOR 2009 There are a number of changes that the department should make for 2009, some of which are listed in the prior section on allegations. But, several critical strategic lessons stand out: 1. Invest in what works, stop funding what does not. This sounds obvious, but is often overlooked. It was clear during research that some programs provided a higher quality experience to young people than others; some programs can demonstrate results and others do not. Yet, the number of organizations that receive funding year-after-year suggests that host organizations are not asked to demonstrate that they prepare young people for jobs later in life. 2. Maintain a commitment to providing quality experiences to a large number of District youth. It was not the findings of this report that aggressive recruitment goals alone caused problems; rather, the more critical root cause was the organization was unable to implement these goals effectively. Fraud and abuse, while important to highlight and eliminate, was not the norm. Rather, the youth that do not bother to look for summer work until June may in fact be the very young people who the city should target and build its operation around. 3. Reassess leadership, skills and people required to manage and run the organization, and put the right people in the right job. The operational complexity of maintaining a 20,000-person summer program offering quality jobs likely requires different skills and knowledge than running a 12,000-person program with an April registration deadline. It demands a commitment to and comfort-level with a sophisticated amount of data management, project tracking, quantifiable program evaluations and technology. A complete reorganization and a top-down assessment of whether individual skills match requirements should occur within the Office of Youth Programs. 4. Enforce time entry requirements at host organizations. Despite clear language in contracts that require time tracking, no incentives exist or are utilized to enforce these requirements. Host organizations not compliant with time entry should be subject to penalties, and consideration for future de-funding. 5. Have job assignment available for youth as they register, with goal of registering a higher percent of participants before the program begins, by securing contracts earlier or migrating to multi-year contracts with providers. The delays that plagued the 2008 program cannot continue in 2009. Given that the District knows that it will need host organizations every year, multi-year contracts provide a vehicle to eliminate (except for 6. Relentlessly challenge assumptions, and train people how to respond to unexpectedly large workflows. Clearly, assumptions for around the timing and progress of preparations need to be thoroughly verified by data. It is not enough to just discuss unexpected events. Contingency plans should be developed. Staff should be trained with processes to adapt to less-than-ideal scenarios, most notably how to build processes and workflows to manage large volumes of in-person requests. 7. Eliminate all paper processes. Reliance on paper-based processes, even when online alternatives were available, is a pervasive theme of this report. The entire workflow, from registration and placement to time entry, needs to be managed online. 8. Continually validate information to stop fraud and abuse before or just after it occurs. The actions taken by OCTO during the fourth pay period proved that it is possible to identify waste, fraud, and abuse quickly to eliminate it from the program. This requires a relentless commitment to searching for data “errors” as they appear in the system. 9. Integrate the debit card with the one card. Each program, individually, was effective in 2008. It remains inefficient, however, to require young people to register for cards and maintain two separate cards. Given the presence of one’s photo on the ID, it could also the city’s ability to guard against abuse. APPENDIX A: TIMELINE OF FIRST 3 PAY PERIODS
1. The data team working on the project is still clarifying this precise numbers. |
Send mail with questions or comments to webmaster@dcwatch.com
Web site copyright ©DCWatch (ISSN 1546-4296)
