The 2018-19 Budget

Evaluating FI$Cal

In this analysis, we describe the Financial Information System for California (FI$Cal) Project and the major changes to the project based on the newly released Special Project Report (SPR). We also describe the Governor’s 2018-19 budget proposal to fund activities associated with the new SPR at the State Controller’s Office. Finally, we make associated findings and recommendations.

Background

FI$Cal Project. Since 2005, the administration has been engaged in the design, development, and implementation of the FI$Cal Project. The project is extremely ambitious and complex, resulting in it being the most costly state information technology (IT) undertaking ever. The FI$Cal Project replaces the state’s aging and decentralized IT financial systems with a new system that will integrate the state’s budgeting, accounting, cash management, and procurement processes. Aside from a small number of departments that were deferred or exempted from the project since its onset, the vast majority of departments will manage their finances through the FI$Cal system. The system will eliminate the need for over 2,500 department-specific applications and enable the state financial systems and workforce to function in an integrated environment. The FI$Cal system will also automate processes that are currently highly manual, improve tracking of statewide expenditures, and standardize the state’s financial practices. The integrated system will be utilized in some way by state departments, the Legislature, and the public—allowing greater transparency of the state’s financial data and management. Such transparency was not possible under the state’s fractured financial management infrastructure. (Refer to our March 10, 2016 report, The 2016-17 Budget: Evaluating FI$Cal for a comprehensive description of the project’s history.)

Partnership of Four Control Agencies Manage Project. The FI$Cal Project is managed by a partnership of four control agencies—Department of Finance (DOF), State Controller’s Office (SCO), State Treasurer’s Office (STO), and the Department of General Services (DGS). These partner agencies have unique constitutional and/or statutory responsibilities over the state processes integrated through FI$Cal—budgeting, accounting, cash management, and procurement. State law mandates these partner agencies to collaborate in the development of the FI$Cal system. Additionally, the project developed a governance plan to guide the relationships among these partners.

Department of FI$Cal Maintains and Operates System. In 2016-17, SPR 6 and budget-related legislation established the Department of FI$Cal to provide the ongoing maintenance and operation function for the FI$Cal system and support services for users of the system. The department is currently providing this support for the functions and departments that have already implemented FI$Cal. Based on the most recent estimates, when the department assumes complete responsibility for maintaining and operating the FI$Cal system in 2019-20, the department’s operating budget is expected to be $73.4 million annually and include 274 permanent positions.

California Department of Technology (CDT) Approves and Oversees Projects. CDT is responsible for reviewing and approving IT project proposals developed by state departments. Once CDT approves a department’s project proposal, its role changes to one of providing project oversight. Specifically, CDT provides an independent review of the project to determine if it is on track to be completed on schedule and within budget, and whether it will achieve the project objectives. As part of this review, CDT routinely reports to departments on issues of concern that it has identified, shares lessons learned from other projects, and recommends risk mitigation and issue resolution strategies. Nevertheless, over time, a project may change in scope or deviate from the schedule and/or cost that was established during the approval process. Any significant changes to the project plan are documented and justified in SPRs. The revised project plans are developed by the department and submitted to CDT for its review and approval. CDT’s approval of an SPR constitutes a new agreement between the department and CDT. This process resets the scope, schedule, and cost from which the project’s progress and performance are assessed. In some cases, projects change considerably and several SPRs are required over the life of the project. CDT—or one of its predecessors—has performed this approval and oversight function through the life of the FI$Cal Project.

FI$Cal Project Plan—SPR 7

In this section we discuss FI$Cal’s latest SPR—SPR 7. Specifically, we first describe the overall status of the FI$Cal Project and discuss the major features of the SPR. Then, we describe in detail the two main components of SPR 7—the implementation of the STO and SCO accounting and cash management functions and the onboarding of departments onto FI$Cal. Finally, we discuss concerns that we have previously raised and that continue in SPR 7. We provide associated findings as we discuss each issue.

Overview of SPR 7

Recent Challenges Triggered New SPR. Since the FI$Cal Project began in 2005, it has changed many times in scope, schedule, and cost from what was initially anticipated. The project was most recently operating under its sixth SPR, which was approved by CDT in February 2016. Challenges have caused the project to fall behind schedule, deviating from SPR 6 significantly enough to trigger the need for a new project plan—SPR 7. In particular, as discussed later, the principal trigger for SPR 7 was the need to replan for the implementation of the STO and SCO accounting and cash management functions that were not ready for deployment on schedule.

SPR 7 Establishes an Arbitrary End Point for Project Completion. The completion of an IT project is determined to be the point in time at which all planned system functions have been implemented. Instead, SPR 7 establishes July 2019 as the project’s end point. The project would end on this date even though some planned functions will not be complete and not all anticipated departments will have transitioned onto FI$Cal. We are not aware of other IT projects defining completion in this manner. As a consequence of this approach, the project will not deliver all of the components the Legislature expected when it authorized FI$Cal. Because of this, the Legislature is likely to receive budget proposals in future fiscal years to complete unfinished work originally envisioned within the scope of the FI$Cal Project.

SPR 7 Significantly Changes Project Scope. SPR 7 significantly changes the scope of the FI$Cal Project in two ways. First, it ends the project before SCO fully transitions onto FI$Cal—assuming that the full transition will happen at a later date. Second, it changes the onboarding approach so that not all departments that were previously anticipated to transition to FI$Cal will actually transition. We discuss each of these components of the SPR in detail later.

SPR 7 Does Not Reflect Total Project Costs. Although SPR 7 makes significant changes to the scope of the FI$Cal Project, it only reflects very minor changes in overall project costs. (We note that these changes are mostly technical in nature and do not necessitate a funding request for FI$Cal in the 2018-19 budget. Instead these cost changes will primarily be covered by project savings from prior years.) However, the Governor’s budget proposes significant resources at SCO in support of activities necessitated by SPR 7. SPR 7 does not reflect a total of $25.6 million (total funds) SCO anticipates spending on SPR 7-related activities through 2021-22. We discuss SCO’s budget proposal below. Figure 1 reflects the project costs incurred to date and the future costs as reflected in SPR 7.

Implementing STO and SCO Functions—the “Integrated Solution”

In the next two sections, we discuss the two main changes that triggered the need for SPR 7. We begin by describing the Integrated Solution, the new approach for deploying the STO and SCO accounting and cash management functions introduced in SPR 7, and provide associated findings.

STO and SCO Financial Management Roles in FI$Cal. Both STO and SCO perform key financial management roles for the state. Among the financial management roles that are most impacted by FI$Cal are both offices’ cash management functions and SCO’s accounting functions. In particular, SCO produces the state’s annual Comprehensive Annual Financial Report (CAFR), a report that informs stakeholders, such as bond rating agencies, of the state’s financial condition and how the state manages public resources.

STO and SCO Accounting and Cash Management Repeatedly Delayed. The STO and SCO functions were originally scheduled to deploy in FI$Cal in July 2015. Implementation of these functions were delayed on multiple occasions by prior SPRs because additional time was needed to develop and test the functions. Despite the previous postponements, these functions were not ready for deploying in July 2017, as anticipated by SPR 6, and in May 2017, the FI$Cal Project again postponed the deployment of these functions. According to the project, the delay was once again necessary because the functions could not be developed and tested on time.

SPR 7 Replans the Release of STO and SCO Functions. In light of the persistent challenges, SPR 7 establishes a new approach for deploying the accounting and cash management functions related to STO and SCO. Under SPR 7, STO will deploy all of its functions in FI$Cal in the first quarter of 2018-19—one year later than anticipated by SPR 6. However, SPR 7 significantly changes the release of SCO functions. Rather than transitioning fully to FI$Cal, as prior SPRs planned, SPR 7 introduces an interim solution, the Integrated Solution.

Under SPR 7’s Integrated Solution SCO Will Run FI$Cal and Legacy Systems in Tandem. Under the Integrated Solution, SCO will run the FI$Cal system and its existing legacy accounting systems in tandem. The Integrated Solution develops interfaces between both FI$Cal and SCO’s legacy systems so that data is entered only once (in either system) but then both systems share the data. This way each system can perform the accounting and cash management functions for the state. Under SPR 7, the Integrated Solution will be deployed incrementally throughout 2018-19.

Integrated Solution Allows SCO to Test and Validate FI$Cal. The Integrated Solution is considered an interim solution to be used until SCO fully transitions onto FI$Cal. The partner agencies agreed to this approach because of SCO’s continued concern regarding the performance and accuracy of the FI$Cal system. SCO has sought additional assurances from the project that the system will work properly. According to the SPR, the Integrated Solution allows additional time for testing and validating the FI$Cal system using reports produced by SCO’s legacy systems before SCO fully transitions.

SCO to Generate CAFR Using FI$Cal and Later Decommission Legacy Systems. During this interim phase, SCO’s legacy systems will continue to serve as the state’s official accounting record. Even while SCO’s legacy system is the accounting record, the FI$Cal system will have access to all of the necessary data and functionality to produce the CAFR. While SPR 7 does not identify a timeline for fully transitioning onto FI$Cal, SCO indicates that it plans to use the FI$Cal system to produce the state’s CAFR for the 2019-20 fiscal year. If the CAFR is produced successfully, SCO anticipates it would decommission its legacy systems beginning in 2022-23.

SPR 7 Removes Key Components of Project Scope. The timeline SCO suggests occurs after the project’s completion date of July 2019. By ending the project before SCO has fully transitioned to FI$Cal, SPR 7 removes significant components of FI$Cal’s project scope. Specifically, SPR 7 removes two key goals from the project’s scope: (1) completely transitioning SCO onto FI$Cal and (2) decommissioning SCO’s legacy systems. As a result, any work to transition SCO or decommission the legacy systems—and potential additional costs to do so—would be conducted at a later date and would not be part of the FI$Cal Project.

SCO’s 2018-19 Budget Proposal for Integrated Solution Is Outside of SPR 7. While the 2018-19 Governor’s Budget does not include a proposal for the Department of FI$Cal for any SPR 7-related activities, the Governor’s budget includes a proposal to support the Integrated Solution at SCO. In support of the Integrated Solution, the Governor’s budget includes 30 positions and $5.4 million in 2018-19, and 38 positions and $6.4 million in 2019-20 and 2020-21. These would decline to 36 positions and $6.1 million in 2021-22 and 2022-23, with ongoing costs of $3.9 million and 30 positions. (We note that the same budget proposal includes resources that SCO will need to operate a different portion of the FI$Cal system irrespective of the Integrated Solution. We have no concern with this aspect of the proposal and exclude the associated cost in the costs we note here.) The administration states that these resources are necessary to: (1) develop and test the new accounting and cash management functions in FI$Cal, (2) assist with the development of the Integrated Solution, and (3) support the maintenance and operations of the FI$Cal Project once the legacy systems are decommissioned. As this is a budget proposal for SCO, it is not reflected as FI$Cal Project costs.

Proposed Provisional Budget Bill Language. The administration also requests provisional budget bill language that gives the DOF the authority to review and approve SCO’s progress in meeting the goals of SPR 7 on a quarterly basis. Under the provisional language, even if the Legislature approves the funds in the Governor’s proposal, they would only be allocated to SCO once DOF has verified the department was making adequate progress. As part of this approval process, the administration would submit a 30-day notice to both houses of the Legislature and the Joint Legislative Budget Committee.

LAO Findings

SCO and STO Integrated Solution Presents Some Risk. The state has experienced considerable challenges implementing IT projects successfully. Given that FI$Cal is extremely ambitious and complex, the risks associated with it are even more acute. This is especially true for the Integrated Solution. The successful implementation of the Integrated Solution depends on multiple interfaces between FI$Cal and SCO’s legacy systems so that they can operate in tandem. Developing the interfaces is technically challenging. Should any of the interfaces operate incorrectly, the state’s accounting and cash management data may not be accurate.

Single Integrated Financial System No Longer Achieved Within Modified Scope of SPR. The principal objective of the FI$Cal Project is to develop a single IT system that integrates all of the state’s financial management functions, including budgeting, accounting, cash management, and procurement. This vision was conveyed to the Legislature through prior SPRs and budget proposals. Ending the FI$Cal Project before SCO completely transitions onto FI$Cal means the original vision of the project will not be fully achieved. While the Integrated Solution allows for the accounting and cash management functions to take place in FI$Cal through interfaces, this is not the same as having a single system. SCO’s reliance on the Integrated Solution will also hinder the state’s ability to achieve some of the cost savings envisioned by the FI$Cal Project, as the state will have to continue to fund the maintenance and operation of SCO’s legacy systems until it completely transitions onto FI$Cal. The state currently maintains and operates SCO’s legacy systems at an annual cost of $1.3 million.

Accountability Mechanism Necessary to Ensure SCO Transitions to FI$Cal . . . It is essential to the original vision of the FI$Cal Project that SCO fully transition to the FI$Cal system and decommission its legacy accounting systems. Pursuant to SPR 7, there is no assurances that SCO will ever fully transition to FI$Cal. In fact, SCO has recently indicated that it may maintain some part of its legacy systems even after it fully transitions to FI$Cal. SCO explains that it could do this either because some functionality will not be fully supported by FI$Cal or to maintain historical data. An accountability mechanism is necessary to ensure the original vision of the FI$Cal Project is achieved.

. . . But Provisional Language Delegates Too Much Control to DOF. We think that oversight of SCO’s transition onto FI$Cal is important. However, we find that giving DOF the unilateral authority to withhold appropriated funding on a quarterly bases is too frequent, delegates too much oversight authority and control to DOF, and may be disruptive to SCO’s progress.

FI$Cal Project Cost Should Include SCO Budget Proposal. The Governor’s budget proposal for SCO anticipates spending a total of $25.6 million ($15.1 million General Fund) through 2021-22 implementing the Integrated Solution that will be used by SCO until it is ready to transition onto FI$Cal. The Integrated Solution was developed in collaboration with various project stakeholders, including the FI$Cal partner agencies—SCO, STO, DGS, and DOF—and CDT. As this is the agreed upon solution to address the challenges with deploying the accounting and cash management functions, these costs should be considered part of the FI$Cal Project, and therefore be included within the total project cost for FI$Cal.

Onboarding Remaining Departments to FI$Cal

In this section we discuss the other major change to the project in SPR 7—the onboarding of the remaining departments onto FI$Cal—and provide associated findings.

SPR 7 Onboards Most Remaining Departments in July 2018. Aside from the SCO and STO accounting and cash management functions that we discussed previously, the system functions envisioned within the scope of FI$Cal are nearly complete. Although the project has delivered the completed portions of the FI$Cal system to 88 mostly small departments, many larger departments are not yet using FI$Cal. Those departments that have transitioned onto the FI$Cal system did so slowly over several releases beginning in July 2013. SPR 7 assumes the project will onboard the remaining departments in July 2018, as envisioned in SPR 6. However, the final release is now larger than originally envisioned in SPR 6. Eleven departments that were not ready to transition onto FI$Cal in July 2017 were shifted to July 2018. (Although four departments that were ready to transition onto FI$Cal early were moved from the 2018 release to the 2017 release.) This final release now includes 62 departments. It will be the largest number of departments onboarding at one time, and includes some of the state’s largest departments with the most complex financial needs. See Figure 2 for a list of the departments in the final release.

SPR 7 Will Not Transition All Departments Onto FI$Cal. The FI$Cal Project originally envisioned that the vast majority of state departments would use the system for their financial management. However, SPR 7 anticipates that “a small number” of the departments in the final release, those listed in Figure 2, will encounter challenges and not transition onto FI$Cal prior to the scheduled completion of the project. Pursuant to SPR 7, any department that does not transition in July 2018 will be removed from the project scope. The project has already identified two departments, the California Department of Justice and the Department of Rehabilitation, that will not transition to FI$Cal in July 2018. SPR 7 now considers these departments “deferred,” and they will transition to FI$Cal at a future, unspecified date. Until these departments are ready to transition onto FI$Cal, they will continue to use their legacy systems. According to CDT, an additional ten departments are experiencing challenges as they prepare to transition to FI$Cal and are at risk of being removed from the FI$Cal Project scope. According to SPR 7, any department transitioning to FI$Cal following the completion of the project in July 2018 would be considered outside the scope of the project. It is our understanding that these departments would have to submit their own requests for funding though the budget process to cover the cost of their transition.

LAO Findings

Onboarding of Departments Will Continue Beyond July 2018, Making it Difficult to Track Total Project Costs. FI$Cal has already identified two departments that will not transition to FI$Cal in July 2018 and ten other departments are experiencing challenges as they prepare to transition. This means that departments will continue to implement FI$Cal after the defined end date of the project. The Legislature is likely to receive budget proposals in future fiscal years to complete the onboarding of departments originally envisioned within the scope of the FI$Cal Project. Because each department that implements FI$Cal after July 2018 will be considered outside the scope of the project, the funding needed for the implementation will be difficult to track. We note that if a critical mass of departments are not prepared to onboard in July 2018, this may signal a serious issue and necessitate a new SPR and an extension of the project schedule until the departments could transition onto FI$Cal.

Remain Concerned With Number and Size of Departments in Final Release. While the project indicates it has drawn lessons from onboarding previous departments that position it to succeed in the final release, we remain concerned by the quantity and complexity of departments that have yet to implement FI$Cal and the compressed timeframe proposed for implementing them. We first raised this concern with SPR 6. At the time, the project acknowledged the magnitude of the work associated with onboarding all of the reaming departments but noted that it would be able to successfully complete the onboarding because it could focus resources towards the onboarding since all functionality would be done. (SPR 6 anticipated the STO and SCO functions would be done in July 2017.) However, SPR 7 adds the complex Integrated Solution and deployment of STO functions to the 2018 release. The significant risk that the project’s resources will be overwhelmed when it tries to implement the July 2018 release is even truer today given the significant expansion of workload associated with the final release.

Previous Concerns That Continue in SPR 7

In the following section we discuss concerns that continue in SPR 7.

Continued Challenges Closing Month- and Year-End Financial Statements Possible. SCO uses the year-end financial statements from departments to prepare the state’s CAFR. While the financial statement functions in FI$Cal are working as intended, some departments already using FI$Cal have experienced difficulty and have failed to close their month- and year-end financial statements on time. Given the difficulty the project has had with smaller departments being unable to close their financial statements on time, it may face challenges in July 2018 when deploying FI$Cal to a greater quantity and more complex departments. If departments are unable to close their year-end financial statements on time, it could affect SCO’s ability to produce the CAFR. Rather than relying on actual data, in prior years the CAFR was developed using estimates for a few small departments that were not able to close their year-end financial statements on time. If some of the larger departments in the July 2018 release were unable to close their year-end financial statements on time, the state would face serious difficulties in developing the 2018-19 CAFR.

Project Failed to Meet Various Conditions Previously Set by CDT. As part of its IT project approval and oversight responsibilities, CDT issues a determination letter following its review of SPRs. In the past, CDT has conditionally approved SPRs. CDT approved SPR 7 in February 2018, but included a number of conditions in its approval letter. Some of the key conditions established by CDT include requiring that FI$Cal provide CDT (1) a contingency plan to address the risks associated with the July 2018 release, (2) a comprehensive schedule that identifies the staff responsible for the activities on the schedule, and (3) a quarterly report on the projects health and progress. In all, CDT’s letter includes ten conditions. However, it is unclear to us whether there will be any consequences if FI$Cal does not meet these conditions. Although FI$Cal eventually met the five conditions established by CDT as part of its approval of SPR 6, it failed to meet the conditions in a timely manner. It does not seem as if FI$Cal faced any consequences for failing to comply with CDT’s conditions for SPR 6 in the timeframe established by CDT. Given FI$Cal’s track record and CDT’s reluctance to hold FI$Cal accountable, we are concerned about FI$Cal’s ability to successfully address the conditions established by CDT in SPR 7.

Staff Vacancy Rate Continues to Present Challenges. Throughout the course of the implementation of FI$Cal, the project has struggled to fill all positions. As of December 2017, the project had an 18 percent vacancy rate (66 of 368 authorized positons unfilled). According to CDT, the project’s vacancy rate ranged from 17 percent to 24 percent during the 2017 calendar year. The high vacancy rate may make it challenging for the project to adequately support the large quantity of departments slated for onboarding in the upcoming release and may jeopardize the aggressive schedule for the Integrated Solution.

Additional SPR May Be Necessary. The FI$Cal Project involves the development of an extremely ambitious IT system and significant work remains before the system is fully implemented. Given the scope of the remaining work and signals from oversight entities that some project activities continue to experience challenges, we think an eighth SPR may be necessary, which would further extend the project schedule and increase costs.

LAO Recommendations

In light of our findings related to SPR 7 and the Governor’s 2018-19 budget proposal for SCO, we make the following recommendations related to FI$Cal.

Approve SCO’s Budget Proposal for 2018-19 . . . The principal objective of the FI$Cal Project is to develop a single IT system that integrates all of the state’s financial management functions, including budgeting, accounting, cash management, and procurement. Successfully transitioning SCO onto FI$Cal is vital to achieving this objective. We understand that the Department of FI$Cal and SCO have determined that the Integrated Solution is a reasonable approach that addresses SCO’s concerns regarding the performance and accuracy of the FI$Cal system while supporting the eventual transition of SCO onto FI$Cal. Therefore, we recommend the Legislature approve SCO’s 2018-19 budget proposal for the Integrated Solution.

. . . But Reject SCO’s Multiyear Funding and Proposed Budget Bill Language. However, we recommend that the Legislature reject SCO’s proposal for 2019-20 and beyond. Doing this would require the administration to make budget requests in future fiscal years. These requests would provide the Legislature with an opportunity to exercise oversight over SCO’s progress on the implementation of the Integrated Solution, the timeline for fully transitioning onto FI$Cal, and the plan for decommissioning the legacy accounting systems. We also recommend the Legislature reject the provisional budget bill language proposed in the Governor’s budget for SCO. We believe the budget bill language is unnecessary given that, under our recommendation, the Legislature would consider funding requests for the Integrated Solution on a year-by-year basis.

Adopt Budget Bill Language Clarifying Project Completion of FI$Cal. Under SPR 7, the costs of fully transitioning SCO onto FI$Cal and onboarding of departments that are postponed following the July 2018 release are considered to be outside of the project. We recommend the Legislature adopt budget bill language that clarifies that the Legislature does not consider the project to be complete until all previously planned functionalities have been implemented and all departments have been onboarded to FI$Cal. Our recommendation would allow for an accurate accounting of all project costs. (As we believe that the Department of FI$Cal should track SCO’s costs for the Integrated Solution as a project cost, we expect FI$Cal to make future budget requests related to the Integrated Solution and decommissioning of SCO’s legacy systems.)