“Cradle to Career” Data System

Summary

State Recently Initiated Planning for an Integrated “Cradle to Career” Data System. In response to longstanding concerns that California lacked an integrated education data system that would allow policymakers, educators, families, and others to answer key questions about student progression and outcomes, the 2019‑20 budget package provided $10 million non‑Proposition 98 General Fund to create such a system. The one‑time funds were provided to the Office of Planning and Research (OPR) for initial planning activities, including forming a multiagency work group tasked with making recommendations on key components of the data system. The work group was to transmit its recommendations to the Legislature through three statutorily required reports. The Legislature received the first work group report on December 31, 2020, with the second and third work group reports anticipated by April 1 and June 30, 2021, respectively. To date, approximately $4 million of the original appropriation has been spent, with $6 million still available.

Governor Proposes to Fund Several Work Group Recommendations. In late January 2021, the Governor notified the Legislature of his intent to authorize the remaining 2019‑20 funds for specific start‑up costs. Also in January, the Governor proposed providing $14.5 million non‑Proposition 98 General Fund ($11.5 million ongoing, $3 million one time) for various other project costs in 2021‑22, including creating a new “Cradle to Career” office within the Government Operations Agency (GovOps). The Governor’s 2021‑22 budget request included an additional $3.8 million ongoing Proposition 98 General Fund for statewide scaling of the California College Guidance Initiative (CCGI), a college planning and advising tool currently used by some school districts. The Governor’s proposals largely reflect the recommendations contained in the work group’s first report. Notably, the Governor proposes to have GovOps manage the system and an 18‑member governing board—primarily consisting of data providers such as the California Department of Education and higher education segments—to oversee the system.

Opportunities to Improve on Governor’s Approach. Though these budget proposals have some meritorious aspects, we have several concerns with them. Our main concerns are the lack of justification for several proposals and the lack of standard project review and oversight, which is typically provided through the California Department of Technology’s (CDT’s) Project Approval Lifecycle (PAL) process. Instead of the Governor’s package of proposals, we recommend the Legislature take a more incremental approach—approving a smaller amount of remaining 2019‑20 funding to hire GovOps staff to begin the PAL process, then wait to approve further budget requests until better cost estimates are available. Additionally, rather than funding CCGI expansion in 2021‑22, we recommend the Legislature provide one‑time funding for an independent evaluation, then use the results of the evaluation to inform the Legislature’s next steps in 2022‑23. Finally, we recommend the Legislature increase the number of public members on the governing board so as to mitigate potentially negative incentives for the data providers to control data decisions and use the system primarily for their own narrower interests.

Introduction

In this post, we (1) provide background on progress to date toward creating an integrated education data system and other related data tools (collectively referred to as the “Cradle to Career” data system), (2) describe the Governor’s proposals to fund the development of specific components of this system, (3) assess these proposals, and (4) make associated recommendations.

Background

In this section, we provide background on the state’s education data systems and recent planning efforts to connect them.

California’s Extensive Education System Lacks Key Data Linkages. California’s education system is made up of numerous segments and other entities. Specifically, the system includes early education programs, elementary and secondary schools, county offices of education, community colleges, and universities in both the public and private sectors. Currently, each of these entities collects and maintains data on its students, but the data generally are not linked across the segments of education (such as from high school to community college). Not linking data limits the ability of policymakers, educators, researchers, parents, and others to get answers to many basic questions about student progression from preschool through K‑12 education, through higher education, and into the workforce.

Some Cross‑Segmental Data Sharing Occurs in California. Although California has no comprehensive integrated data system, some cross‑segmental data efforts exist. Among the most notable of these efforts are:

• CCGI. About 100 school districts participate in CCGI, which a nonprofit entity administers. The state currently funds CCGI as part of the California Department of Education’s (CDE) budget. Partner districts can upload verified academic transcript data into students’ accounts on CaliforniaColleges.edu. When students from these partner districts apply to a California Community College (CCC) or California State University (CSU), certain high school data is shared. The college or university, in turn, can use the data to inform decisions about admissions and course placement.
• eTranscript California. This platform transfers student‑level course, grade, and certain other data files (such as degrees earned) electronically across California’s higher education segments. Most of these electronic transcripts are for community college students applying for transfer admission to CSU or the University of California (UC). Using funds provided by the state, the CCC Chancellor’s Office contracts with the CCC Technology Center (located at Butte College) to manage and operate the platform.
• Cal‑Pass Plus. Most school districts, every community college, and some public and private four‑year institutions voluntarily participate in Cal‑PASS Plus, a project overseen by the CCC Chancellor’s Office. Cal‑PASS Plus links data from participating institutions, then analyzes the data. It provides its findings to participating institutions—sharing outcomes with them so they may learn more about their students and how to improve their outcomes.
• Employment Outcomes. Each public higher education segment has an agreement with the Employment Development Department (EDD) that allows it to identify the quarterly earnings of its graduates. The data are matched using social security numbers, which most students provide when they apply to college.

State Recently Initiated Planning for an Integrated Cradle to Career Data System. As part of the 2019‑20 budget package, Chapter 51 of 2019 (SB 75, Committee on Budget and Fiscal Review) provided $10 million one‑time non‑Proposition 98 General Fund to the OPR for initial work related to developing an integrated data system. The budget package included intent language that the data system be built to “advance academic and governmental research on improving policies from birth through career” as well as “create direct support tools for teachers, parents, advisors, and students.” Of the $10 million appropriation, about $4 million was for work group planning and related activities, with the remaining approximately $6 million available for initial development of the data system. The development funds are contingent on (1) the Legislature’s receipt of three work group reports (discussed in the next paragraph), (2) approval of an expenditure plan by the Department of Finance (DOF), and (3) notification to the Joint Legislative Budget Committee (JLBC). Chapter 51 identified a total of 17 representatives from 15 specified state entities to be in the work group. The director of OPR was charged with contracting with an outside “planning facilitator” to support the group.

Three Planning Documents Required to Be Submitted to the Legislature. The budget package required the planning facilitator to submit three reports to the Legislature and DOF that detail the work group’s advice on matters such as the type of data system to create and the entity to operate the system. Figure 1 highlights key information to be included in each report. Though the Cradle to Career data system is an information technology (IT) project and the CDT was designated as a member of the work group, Chapter 51 did not require OPR to plan the data system through the CDT’s Project Approval Lifecycle (PAL) process. The box below describes the state’s PAL and project oversight processes.

Work Group Recommends Three Cradle to Career Data Projects. The planning facilitator—which OPR selected to be WestEd, a nonprofit organization based in San Francisco—convened the first work group planning meeting in January 2020. The Legislature received the first required work group report on December 31, 2020. The report was delayed by six months due to scheduling disruptions resulting from the pandemic. To fulfill the requirements of Chapter 51, the work group recommends the state fund three data projects: (1) an integrated education data system (referred to in the report as a “P20W” data system) that links records from various state agencies, including CDE, the higher education segments, the Commission on Teacher Credentialing, EDD, and the Department of Social Services; (2) an expansion of CCGI to school districts throughout the state; and (3) an upgrade to eTranscript California to, among other things, include specific skills students acquired through competency‑based education or other forms of nontraditional learning. The report recommends rolling out these projects over a five‑year period, starting with the integrated education data system and CCGI expansion project.

Work Group Also Recommends a Governance Structure... The work group also recommends that the data system be overseen by a governing board made up of 18 members, consisting of: (1) 12 agencies contributing data to the integrated education data system and (2) 6 public members—4 appointed by the Governor and 2 appointed by the Legislature. The work group recommends that decisions by the governing board be made by two‑thirds vote. Figure 2 shows the proposed members of the governing board. The work group also recommends two advisory boards—the Data and Tools Advisory Board and the Community Engagement Advisory Board. Whereas the first advisory board would examine whether the data system is providing actionable information, the second advisory board would examine whether the public is aware of the data system and knows how to use the information in it.

...And a Managing Entity. The governing board would be charged with hiring an executive director, who would establish and operate a “Cradle to Career” office within the state’s GovOps. GovOps would build and operate the data system, with an evaluation after five years to determine if it or another entity should have that role. The box below provides an overview of GovOps.

Legislature Is Likely to Receive Remaining Work Group Reports Later This Year. OPR indicates that the work group’s required interim report on the status and progress of the final report will be provided to the Legislature by April 1, 2021. OPR indicates that the third and final report, which will include an estimate of ongoing costs for the data system and recommendations on issues such as protecting privacy and security, will be submitted to the Legislature by June 30, 2021.

Governor’s Proposals

In this section, we describe the Governor’s proposals to fund the work group’s recommendations.

Closely Follows Work Group’s Recommendations. The Governor’s trailer bill language largely reflects the recommendations in the work group’s first report, including the composition of the governing board, the creation of two advisory groups, and identification of GovOps as the managing entity to incubate the Cradle to Career project. The trailer language generally defers to the governing board to determine a specific time line for implementing Cradle to Career projects.

Provides Both Ongoing and One‑Time Funds to Begin Development and Implementation Phase. Specifically, the Governor’s budget includes:

• $14.5 million non‑Proposition 98 General Fund ($11.5 million ongoing, $3 million one time) to GovOps. A portion of the funds would support 12 staff (including an executive director) in 2021‑22 at a newly created Cradle to Career office within GovOps. (The administration proposes increasing staff to 16 in 2022‑23 and providing an additional $500,000 ongoing funding for GovOps at that time.) The remaining funds in 2021‑22 would be used to cover various operating and technology acquisition costs related to the integrated data system, including funds to upgrade CDE’s K‑12 database.
• $3.8 million ongoing Proposition 98 General Fund to increase the number of CCGI partner school districts. To support statewide expansion, funding is also provided to allow CCGI to receive data directly from CDE’s K‑12 database. (CCGI currently receives data directly from each partner school district.)

Notifies JLBC of Intent to Spend Remaining 2019‑20 Funds. In addition, in January 2021, DOF sent notification to JLBC of the administration’s intent to spend the remaining $6 million from the 2019‑20 appropriation for initial costs related to the integrated education data system and CCGI expansion. Specifically, the administration would begin to use the funds in the current year to (1) hire interim staff at GovOps, (2) commence a pilot (proof of concept) for the integrated education data system, (3) begin to scale CCGI (primarily in the Central Valley), and (4) support various one‑time technology and software acquisition costs. In February 2021, DOF agreed to give JLBC until April 15, 2021 to decide whether to concur with this expenditure request.

2021‑22 Budget Proposal Could Change at May Revision. As mentioned earlier, the work group’s final legislative report is required to include estimates of ongoing costs for its recommended Cradle to Career data system projects (consisting of the integrated education data system as well as CCGI and eTranscript California expansions). To the extent such estimates are available in time (such as in the April 1, 2021 interim report), DOF has indicated it would consider revisions to the Governor’s budget proposals as part of the May Revision.

Assessment

In this section, we provide our assessment of the Governor’s proposals (for both current‑ and budget‑year spending). While we agree with the Governor that having an integrated education data system would be an asset for the state and GovOps is a reasonable managing entity for that system, we have a number of concerns with the administration’s proposed approach. Below, we identify the meritorious aspects of the Governor’s proposals, then turn to our concerns.

An Integrated Education Data System Could Have Significant State‑ and Local‑Level Benefits. The administration’s goal to create an integrated education data system for California is in line with a longstanding priority of the Legislature. Such a data system could give the Legislature a more holistic view of the state’s education system and allow policymakers to make more informed budget and policy decisions. An integrated data system also could provide more information to educators about what happens to their students—disaggregated by gender, racial/ethnic group, and other demographics—after leaving their particular education segment, thereby providing greater insight into the effectiveness of current practices. Figure 3 provides examples of the kinds of crosscutting policy questions the Legislature, educators, and others could answer with an integrated data system.

Tasking GovOps With Managing System Could Further State Synergies. GovOps does not have any particular experience or expertise with developing and implementing data systems. Charging GovOps with managing the Cradle to Career data system, however, does have some potential advantages. In particular, GovOps is working to improve how the state collects, manages, and uses data through its Statewide Data Strategy. This initiative, which is led by the state’s Chief Data Officer, is focused on aligning other integrated data system development efforts in the state, standardizing interagency data exchange agreements, and developing a statewide approach to data literacy and data skills training. We believe opportunities exist to create synergies between these existing state efforts and the proposed education data system. A new office within GovOps thus could be a reasonable place to incubate the Cradle to Career data system.

Several Concerns With Project Planning and Oversight Processes. CDT’s PAL process was designed to improve the quality of IT projects and provide the Legislature with a more complete plan and set of cost estimates before funding the project. As noted earlier, the Cradle to Career data system currently is not being planned through PAL. The administration has indicated that it expects the integrated education data system project to use the PAL process, but the administration is unclear at what point the project will enter that process. In addition, the administration is not clear whether there will be independent project oversight and independent verification and validation of the project, which also are typical for state IT projects. Furthermore, the administration has indicated it does not expect CCGI or eTranscript California to go through the PAL process. The administration asserts that given CCGI and eTranscript California are managed by nonstate entities that would enter into a contractual relationship with the Cradle to Career governing board, those projects are not subject to the PAL process. Given the administration identifies potential data flows and information sharing between the three projects, however, we believe CCGI and eTranscript California also should go through the PAL process.

Proposed Budget Makes Number of Unsubstantiated Assumptions, Lacks Basic Details. The Governor’s estimates of one‑time and ongoing project costs use a number of assumptions that are based on incomplete information. Typically, agencies undertake market research to identify what technical solutions are available for a project and determine how much it would cost to implement those solutions consistent with a project’s system requirements. The administration conducted some market research to estimate some one‑time costs, but other one‑time and ongoing project costs appear to be rough estimates based on one other state data system. For example, the administration indicates that some of its estimates were based on what it cost the state to develop a data system on homelessness, scaled up to take into account the much larger scope of the integrated education data system. This approach to estimating costs is not standard practice and can result in inaccurate cost estimates. Other data systems can have significantly different technical requirements resulting in different project solutions that have much higher or lower costs than the proposed project. Also a departure from standard state budget practices, the Governor’s budget does not provide workload justification for the requested GovOps positions. The administration indicates that it expects the work group to provide workload justification later in the spring. Much of this basic cost and staffing information, however, would have been made available to the Legislature had the work group used CDT’s PAL process.

Premature to Scale CCGI Without Evaluation. CCGI has a number of features designed to help students, parents, and counselors. For example, CCGI provides students with information about the college and financial aid application process, and it provides an education planner that allows students and high school counselors in partner districts to track progress toward meeting CSU and UC course eligibility requirements. Though these types of features have the potential to be beneficial, to date, the Legislature has not provided funding specifically for an independent, comprehensive evaluation and CCGI has not used its base funding for such an evaluation. As a result, the Legislature lacks basic information on CCGI’s usage and outcomes. For example, the Legislature cannot answer the following questions:

• To what extent are CCGI’s various features currently used by students, schools, community colleges, and universities?
• Which particular features or services do they find most valuable?
• If certain features are not widely used, why not? Could modifications be made to CCGI so these features are more useful for students and other potential users?
• What are the reasons why non‑partner school districts have chosen not to participate in CCGI to date?
• What is the impact of CCGI on student outcomes? For example, to what extent are students who use CCGI more likely to enroll in college than their peers who do not use CCGI? To what extent do differences exist among students by gender or from certain racial/ethnic groups?

The administration has indicated that while its current proposal is to provide $3.8 million ongoing for CCGI, at full implementation ongoing costs could be about $20 million annually. (The final legislative report is expected to include updated cost estimates for the project.) We believe the Legislature should have answers to the above types of questions before embarking on a costly expansion of CCGI.

Insufficient Role for Non‑Data Providers and Legislature on Governing Board. While proposing to reserve one‑third of governing board seats for public members is a good start, we believe having greater public representation would have certain benefits. Most notably, having more public representation on the governing board is an important way to mitigate the possibility of data providers using the system primarily to further segment‑specific interests. This is a particularly important issue given that the Governor’s proposed trailer bill language tasks the governing board with developing the data‑request process used by education researchers and other members of the public. Historically, certain segments have been reluctant to approve such requests and share data. We also believe the Legislature should have a larger role in selecting members of the governing board. Under the Governor’s proposal, the Governor would get to select four public members while the Senate and Assembly combined would get to select only two public members.

Recommendations

In this section, we make several recommendations designed to improve on the Governor’s proposals.

Require Administration to Use Standard Approval and Oversight Processes. Given the state’s mixed track record with IT projects, we recommend the Legislature specify in statute (such as through trailer bill language) that the administration use the PAL process to plan the entire Cradle to Career data system (including CCGI and eTranscript California), as well as have independent oversight, verification, and validation of the system upon completing the PAL process. We note that the administration can submit the project through the PAL process now, and CDT can use the project’s prior planning efforts to help it decide at what stage of the PAL process the project should begin. The project therefore would not be significantly delayed because of the requirement to use the PAL process. Furthermore, CDT could determine that CCGI and eTranscipt California are largely updates to existing IT systems that are low risk. In this case, CDT could delegate those projects back to their respective administering entities to complete the planning phase. If the Legislature subsequently approved and funded those projects, they then could proceed with development and implementation without needing to complete the rest of the PAL process.

Require Administration to Provide More Justification for Budget Requests. We also recommend the Legislature request that the administration provide additional information that would support its budget proposals. At a minimum, we recommend the Legislature request: (1) market research that evaluates available technical solutions for the system and estimates the one‑time and ongoing project costs based on system requirements, (2) a schedule for completion of project activities, and (3) workload justifications for the requested GovOps positions. In the absence of this information, we recommend the Legislature wait to approve the Governor’s budget request until the project completes at least Stage 2 (the Alternatives Analysis) of the PAL process. In the meantime, the Legislature could consider approving some amount of remaining 2019‑20 funding for GovOps to hire the staff necessary to begin the PAL process. The next paragraph provides an illustration of what an alternative 2021‑22 budget might look under our recommended approach.

LAO Budget Alternative Funds Some Staff and Start‑Up Activities While Awaiting Better Information. Figure 4 compares the Governor’s budget proposals with an LAO budget alternative. Our alternative would designate $1.2 million in remaining 2019‑20 funds for board meetings and some GovOps project staff as well as knowledge transfer activities from WestEd to GovOps. (As discussed in the next paragraph, our LAO alternative also includes a small amount of 2019‑20 funds for CCGI.) At this time, the alternative would not fund the costs of the pilot or technology procurement activities, as these components would be undergoing further analysis as part of the PAL process. Upon completing the PAL process, better estimates of these costs, as well as other potential project costs, would be available. We recommend the administration submit a future budget request on these project components once more reliable cost estimates are available.

Fund CCGI Evaluation. Instead of providing funding for CCGI expansion in 2021‑22, we recommend the Legislature provide one‑time funding for an evaluation and then use the results of the evaluation to inform the Legislature’s next steps in 2022‑23. Based on our discussions with evaluation firms, we believe that about $250,000 would be sufficient for such a study. The study could commence some research activities in summer 2021, with other activities (especially those involving student feedback) commencing in fall 2021, with findings provided to the Legislature by spring 2022.

Increase Public Representation on Governing Board. Finally, we recommend the Legislature increase the number of non‑data providers (public representatives) on the governing board so they are at least equal to—if not greater than—the number of seats held by the data providers. We further recommend specifying that the public members include representatives from the administration as well as bipartisan representation from the Legislature.