The 2019-20 Budget: Creating an Integrated Education Data System

The 2019-20 Budget: Creating an Integrated Education Data System GABRIEL PETEK LEGISLATIVE ANALYST MAY 6, 2019 Summary Creating an Integrated Education Data System Has Been a Key Legislative Priority. California’s education system is made up of numerous segments. Currently, each segment collects and maintains data on its students but the data generally are not linked across the segments. This limits the ability of policymakers, educators, researchers, parents, and others to get answers to many basic questions about student progression from early education through K-12 education, through higher education, and into the workforce. In recent years, the Legislature has commissioned intersegmental work groups to study and make recommendations on developing an integrated data system. Legislation resulting from the work groups’ efforts was vetoed by the Brown Administration, citing cost and other concerns. Recommend Legislature Build on Governor’s Proposal and Prior Efforts. The Governor proposes $10 million one time for the development of an integrated data system. Of that amount, $3.1 million would fund a new work group to study and make recommendations to the administration on the governance, data structure, and other elements of an integrated system. The remaining $6.9 million would be for data matching and implementation of the project ultimately approved by the administration. We believe the Governor’s proposal has some positive components, but also some shortcomings. We recommend the Legislature adopt an alternative approach. Compared with the Governor’s proposal, our recommended approach leverages planning that has already been done, uses available funding more effectively, and strengthens oversight of the project throughout the development and implementation process. In this brief, we provide (1) background on the state’s for implementing state technology projects, including education data systems and past efforts to connect data systems. them, (2) describe the Governor’s proposal to develop California’s Public Education System Is Comprised an integrated education data system, (3) assess the of Many Entities. Preschool in California is offered by proposal, and (4) make associated recommendations. many local entities, including schools and nonprofit organizations. The California Department of Education Background (CDE) administers annual contracts for these providers In this section, we provide an overview of California’s to operate. Kindergarten through grade 12 (K-12) is run education structure and education data systems; by roughly 1,000 school districts, with elected boards discuss past legislative efforts to create an integrated governing at the local level and CDE providing support education data system; and describe required processes at the state level. CDE reports to the independently elected State Superintendent of Public Instruction. The analysis full gutter 2019-20 BUDGET State Board of Education (SBE), with members two-thirds of states with integrated data systems, appointed by the Governor, interacts regularly with education and workforce agencies send a copy of CDE in guiding implementation of state education student-level and wage records to a designated laws. Public higher education consists of three central repository (“data warehouse”) on a regular segments—the California Community Colleges schedule (such as at the end of each academic (CCC), California State University (CSU), and term). Staff then match the data records and assign University of California (UC). UC and CSU have a nonpersonally identifiable P-20 number to each system governing boards that oversee their 10 student. The integrated data are then available to and 23 campuses, respectively. The CCC also perform analyses and run reports. By comparison, has a system governing board, but its autonomy other states have developed “federated” data is more limited, with each of the system’s systems, whereby each education and workforce 73 community college districts having its own local entity collects and keeps its own respective data governing board. In addition to all these entities, (as opposed to copying and sending it to another California’s public education system includes other entity). Under the federated model, separate data agencies—including the Commission on Teacher systems connect to each other upon request. Credentialing (CTC), which is responsible for the Each time a query is required to be performed for credentialing of K-12 teachers in California, and the a report or other purpose, each segment “opens California Student Aid Commission (CSAC), a state up” its database for a short period of time—long agency responsible for administering financial aid enough for the data matching and analysis to programs. occur—at which point the segments lock up State’s Education Data Systems Are Siloed. their data systems again. Federated systems Education data in California currently are siloed, typically have developed in states in which entities being maintained and managed in several separate contributing data were unable to agree on the data systems. Most notably, CDE collects K-12 location and control of a centralized data system. data (including student demographics, courses Both types of data systems report aggregated taken, grades earned, and high school graduation). information (averages and group statistics), which, For tracking and reporting purposes, CDE assigns for privacy purposes, is stripped of any personally to each K-12 student a unique, nonpersonally identifiable information. identifiable number (known as a “statewide Some Cross-Segmental Data Sharing student identifier”). Each of the public higher Occurs in California. Although California has education segments collects and maintains its no comprehensive P-20 data system, some own student-level data and uses different student cross-segmental data efforts exist. Among the most identifiers from CDE. State-subsidized early notable of these efforts are: education programs (including most preschool • Most school districts, every community programs) are part of a separate data system that college, and some CSU, UC, and private is not connected with CDE’s student data system four-year institutions voluntarily participate and generally does not assign statewide student in Cal-PASS Plus, a project overseen by the identifiers. Given that existing data systems are CCC Chancellor’s Office. Cal-PASS Plus siloed, tracking and assessing students’ progress collects and links data from participating across segments is difficult. institutions, then analyzes the data. It California Is One of Only a Handful of States provides ongoing dashboards and reports Without Some Type of Integrated Education to participating institutions—confidentially Data System. According to the Education sharing outcomes with them so they may Commission of the States, as of November learn more about their students and how to 2016, California was one of only eight states improve their outcomes. neither possessing nor in the process of creating • Thirty-nine school districts participate in the an integrated education data system. In about California College Guidance Initiative, which 2 LEGISLATIVE ANALYST’S OFFICE analysis full gutter 2019-20 BUDGET is housed within the Foundation for California of such a data system. The two work groups Community Colleges. School districts included many education representatives, including upload verified academic transcript data into from the Superintendent of Public Instruction, the students’ accounts on CaliforniaColleges.edu. CCC Chancellor’s Office, the CSU Chancellor’s When students from participating districts Office, and the UC Office of the President. The apply to CCC or CSU, certain high school governance work group also included legislative data is shared, with colleges being able to use staff from both houses and both parties as well as the transcript data to inform decisions about the Department of Finance. In addition, governance admissions, course placement, and student work group members invited the participation of services. an advisory group, which included teacher, school • Each public higher education segment has an administrator, and school board associations; agreement with the Employment Development CTC; and other organizations, associations, and Department (EDD) that allows it to identify the government entities. quarterly earnings of its graduates. The data SB 1298 Technical Work Group Identified are matched using social security numbers, Roadmap to Creating Integrated Education Data which most students provide when they apply System. The technical work group met throughout to college. 2009 and completed its report in early 2010. The report identified a series of recommended steps A Former State Agency Previously Housed to undertake over a two-year period to build what Higher Education Data. While it is now could become either a centralized or federated defunct, the California Postsecondary Education data system. These identified steps included Commission (CPEC) formerly collected and (1) identifying the key questions to be answered by maintained data from the three public higher the integrated data system; (2) creating standard education segments to support its statutory data definitions across the segments; (3) assigning research and advisory responsibilities. CPEC’s a unique identifier to each student in the P-20 data system collected information such as system; (4) establishing policies and procedures to transfers between CCC and universities, degrees ensure security and privacy of student data; and awarded across the three segments, and statewide (5) creating a single portal and data tools (such enrollment projections. When then-Governor Brown as dashboards and reports) for policymakers, vetoed funding for CPEC in 2011, the agency’s educators, researchers, and the public to access data system was shuttered. the data. The work group also recommended that Statute Created Two Work Groups to Advise the data system be built with the capability to Legislature on Creating Integrated Education eventually link education records with records in Data System. For years, the Legislature has health, social services, criminal justice, and other expressed a strong desire to enhance its state data systems. oversight role of California’s education system and SB 1298 Governance Work Group Presented facilitate continuous improvement in education Three Options. The governance work group by developing a P-20 data system. In an effort submitted its report in late 2009. The work group to spur the development of such a system, the envisioned creating a centralized data warehouse Legislature enacted Chapter 561 of 2008 (SB 1298, and recommended that it be governed by a single Simitian). Chapter 561 created two work groups entity whose mission crosscuts K-12 and higher (referred to as “SB 1298 work groups”). The first education. The group recommended that this work group, led by the state’s Chief Information entity include representation from each of the Officer, was charged with identifying the technical contributing education segments as well as a steps toward implementing an integrated education “significant majority” of state-level, non-segmental data system. The second work group was charged members (including bipartisan representation with making recommendations on the governance from the Assembly and Senate and staff from www.lao.ca.gov 3 analysis full gutter 2019-20 BUDGET the administration). Work group members project monitoring reports, CDT identifies issues recommended arranging governance this way so of concern, shares lessons learned from other as to mitigate potentially negative incentives for the projects, and recommends strategies to reduce segments to control data and use it primarily for project risks and fix identified issues. Current law segment-specific interests. The group did not settle generally requires CDE and the CCC Chancellor’s on a particular structure for the new entity but put Office to use the CDT planning process for IT forward three options—creating a joint powers projects costing more than $1 million. CSU and UC authority (JPA), creating a new state agency, or are not required to go through CDT’s process. housing the data system within an existing state Governor’s Proposal agency. Legislature Authorized Creation of JPA In this section, we describe the Governor’s but Governor Vetoed Bill. In response to the proposal to study and fund an integrated data governance work group’s report, the Legislature system. passed SB 885 (Simitian) in 2012. SB 885 Declares Intent to Create Integrated Data authorized the creation of an intersegmental JPA to System. In proposed trailer bill language, the plan, implement, and manage a P-20 data system. administration indicates the integrated system In addition, SB 885 called on the JPA to plan the would be intended to benefit a number of groups, new data system, with the intention of eventually including policymakers, teachers, student advisors, incorporating data from noneducation sources, parents, students, and researchers, as well as including employment, health, and corrections health and human services providers. The proposed agencies. Though SB 885 received nearly trailer bill language gives significant discretion to a unanimous support from legislators, then-Governor work group to determine what such a data system Brown vetoed the legislation, citing, among other might look like. For example, the data system reasons, concerns about cost and the state’s could be centralized or federated. According to “current fiscal constraints.” Since that time, the the Department of Finance, the administration’s Legislature has introduced various bills to create long-term goal would be to expand the data system some type of integrated data system, with those so that educators and other staff across segments bills either failing to make it out of committee or and agencies could access student data. Under being vetoed by Governor Brown. such a vision, the data system could be used as California Has Process for Approving a tool to provide academic advice and connect Information Technology (IT) Projects. Led by students to health and other services. the state’s Chief Information Officer, the California Authorizes State Board of Education to Lead Department of Technology (CDT) is responsible Data System Work Group. As Figure 1 shows, for reviewing and approving IT project proposals the proposed work group would consist of ten developed by most state departments. CDT has education and other departmental agencies, with multiple stages to its project approval process. a total of at least 11 representatives. (CDE would Each stage requires departments to conduct have at least two representatives—one from its specific planning-related analyses and submit an early learning and care division and the other from associated planning document to CDT. Collectively, its analytical and reporting division.) The Governor the planning documents from these stages create a would select the members of the work group comprehensive plan for implementing a proposed IT based on nominations provided by each entity. project. Once CDT approves a department’s project The executive director of the SBE would convene proposal, CDT’s role typically changes to providing and lead the work group. With the approval of project oversight. Specifically, CDT provides SBE, CDE would contract with third party entities independent review of the project—monitoring (referred to as “planning facilitators”) to staff the whether it remains within budget, on schedule, and work group. on track to achieve its established objectives. In its 4 LEGISLATIVE ANALYST’S OFFICE analysis full gutter 2019-20 BUDGET Work Group Charged With Figure 1 Making Recommendations to Governor’s Proposal Would Create Work Group Administration on Data System. Consisting of Ten Entities Under the proposed trailer bill language, the work group would be • State Board of Education (leader) charged with studying and making • California Department of Education a number of recommendations. • California Community Colleges Chancellor’s Office The planning facilitators would • California State University Chancellor’s Office compile these findings and • University of California Office of the President recommendations into two required • California Commission on Teacher Credentialing • California Student Aid Commission reports: • California Labor and Workforce Agency • The first report, due by • Employment Development Department • California Health and Human Services Agency March 2020 (or later if approved by the Department of Finance), is to include CSU Chancellor’s Offices (and requests the UC the work group’s advice on matters such Office of the President) to perform two activities as (1) the type of data system to create, within certain time frames. By December 2019, (2) the data elements that would receive the these segments must “develop a means” of highest priority to be included in the data connecting student records for K-12 students at system, (3) the entity that would manage and the time of their enrollment in CCC, CSU, or UC. secure the data system, (4) the protocols for By December 2020, the planning facilitators are to protecting student privacy and addressing ensure that the same segments begin implementing data security risks, and (5) the entities or this data linkage. persons that would have access to the data. Provides $10 Million One-Time • The second report, due by September 2020 Non-Proposition 98 General Fund for Data (or later if approved by the Department Efforts. These funds would be appropriated to of Finance), is to summarize the work CDE for allocation purposes. Of the total amount, group’s recommendations for “expanded $2 million would be available for CDE to contract and enhanced data system functionality,” with work group planning facilitators. In addition, including (1) plans to expand the data system CDE would receive $200,000 to cover any costs to “incorporate workforce, financial aid, and associated with its planning activities. The other health and human services data,” (2) steps nine entities would receive $100,000 each for to increase data quality provided by each work-group planning activities. In addition, each entity contributing to the data system, and of the four public education segments would (3) a proposed timeline and budget for receive $50,000 to develop a means of matching implementing its recommendations. student data. The remaining $6.7 million would be available to implement the initial phases of the The work group would be required to submit both data system or for the subsequent “expansion reports to the Department of Finance. According and enhancements” phase, contingent on the to the Department of Finance, the administration Department of Finance’s approval. The proposed would then decide how best to respond to the work trailer bill language does not require the project to group’s recommendations. go through any part of CDT’s review or approval Requires Public Education Segments to process. Develop Way to Match and Share Student Assessment Data. Regardless of the work group’s ultimate recommendations, the proposed trailer bill In this section, we provide our assessment of language requires CDE as well as the CCC and the Governor’s proposal. While we agree with the www.lao.ca.gov 5 analysis full gutter 2019-20 BUDGET Governor that having an integrated data system system in place could help the Legislature better would be an asset for the state, we have a number exercise its oversight role and better assess the of concerns with his proposed approach to effectiveness of its policies. Figure 2 provides developing one. examples of the kinds of crosscutting policy An Integrated Education Data System questions the Legislature could answer with an Could Have Significant State- and Local-Level integrated data system. Such a data system could Benefits. In our view, California could benefit also help local educators evaluate and improve their significantly from an integrated education data practices and provide useful, timely information to system. Each year, the state provides billions of students and families. dollars for education. Over the years, it also has Governor’s Proposed Approach Minimizes made major policy changes. For example, over Legislature’s Role . . . Though an integrated data the past decade, the state has implemented the system could provide notable benefits, we have Local Control Funding Formula with an aim toward serious concerns with the Governor’s proposed improving students’ college and career readiness, role for the Legislature in developing the system. funded several new career technical education Under the proposed trailer bill language, the only programs intended to improve the alignment of responsibility for the Legislature would be to high school and college courses, and streamlined appropriate the $10 million in one-time funds. All transfer pathways for CCC students to enroll in other core responsibilities—including choosing the upper-division coursework at the universities. work group representatives, reviewing work group Lacking an integrated data system, the Legislature reports, deciding on the governance structure, has had only a limited ability to assess the impact selecting the data system structure, and approving of those changes. Having an integrated data $6.7 million in funds for project implementation— Figure 2 Questions That Could Be Answered With an Integrated Education Data System 1. Which early education programs and services have the greatest effect on reading and comprehension in elementary school? 2. What are the demographic, program, and course-taking profiles of K-12 students who enroll or do not enroll in postsecondary education? 3. What are the characteristics and educational paths of students who drop out of high school but eventually enroll in a postsecondary institution? 4. Is how districts use their supplemental grants under the Local Control Funding Formula affecting the proportion of their low-income students who enroll in and graduate from college? 5. What are the postsecondary enrollment and completion patterns of students in high school career technical education (CTE) pathway programs compared with similar students not in a CTE pathway? 6. Does dual (concurrent) enrollment by high school students in college courses promote more timely and efficient completion of associate and bachelor’s degree programs? 7. Do students who earn an associate degree for transfer (ADT) at a community college end up taking fewer total units to earn a bachelor’s degree than students who transfer without an ADT? 8. Are students receiving Cal Grant competitive awards more likely to enroll and graduate from college than those eligible students who just missed the cut-off for getting awards? 9. What are the employment outcomes of graduates from CSU and UC teacher preparation programs? 10. Which health and social service programs are most closely associated with improved educational outcomes of K-12 and college students? 6 LEGISLATIVE ANALYST’S OFFICE analysis full gutter 2019-20 BUDGET would fall exclusively to the Governor or managing the new system. By having a majority Department of Finance. In effect, the Legislature of JPA representation come from outside the would be appropriating funds for a project without education segments—as also recommended knowing what it would get in return and without by the SB 1298 work group—the state could any assurance that the final product would be achieve the benefit of ensuring the segments do consistent with its priorities. not limit access to data or use the data for largely . . . As Well as Sidelines CDT From Planning, self-serving purposes. A JPA has the added Development, and Implementation Process. potential benefit of leveraging some administrative Equally as troubling is the lack of any planning and and technology infrastructure already in place at oversight role for CDT in the proposed project. the segments—resulting in possible efficiencies and According to the Department of Finance, the the ability to organize a governing entity relatively project would not include CDT’s involvement due to quickly. the administration’s desire to move quickly. Given A Centralized Data Warehouse Has the state’s historically mixed track record with IT Advantages Over Federated System. Centralized projects, we question the prudence of bypassing data systems are the most common among states CDT’s planning and oversight functions. Completing and have distinct advantages over federated CDT’s project approval process could improve the systems. A federated system requires the segments quality of the project and provide the Legislature to perform a data match every time a query with a more complete plan before funding the is made—requiring them to respond to every project. CDT’s independent oversight also could individual data request. With a data warehouse, highlight risks and alert stakeholders (including the once the data are uploaded and matched, the data Legislature) to any significant changes in the cost, are available for staff to respond to any number schedule, and scope of the project. of queries simultaneously. As a result, centralized State Has Already Given Much Thought systems tend to be more efficient—enabling access to Creating an Integrated Data System. Over to more quickly produced data analyses. the years, the Legislature, education segments, Governor’s Incremental Approach to and related entities have deliberated at length on Adding Data Sources Has Merit. We believe creating an integrated data system. As a result of the Governor’s proposed order of onboarding the these efforts, the segments already have produced segments and other data contributors generally is a technical roadmap for creating such a system. reasonable. Recognizing that all IT projects entail Moreover, the Legislature even passed legislation risks, first connecting K-12 data with data from the authorizing the creation of a JPA for purposes of state’s three public higher education segments, governing a new data system. Given this prior as proposed by the Governor, likely would be work, we are concerned with the administration’s the most straightforward technical process, with approach of effectively starting anew and considerable added value. The proposed trailer bill authorizing another work group. We also are language is not explicit on whether early education concerned with providing $3.1 million for planning data and EDD data should be included as part of facilitators and the new work group to study and the first phase. Given the generally fragmented make recommendations on many of the same state of current early education data, adding child fundamental issues as former work groups. care and preschool data in the first phase seems JPA Governing Model Remains a Promising premature. Because the higher education segments Approach. As noted by the SB 1298 governance already match student records with EDD, including work group, a JPA has several advantages that EDD wage data in the initial roll out, however, could help ensure the successful implementation of likely would be feasible. We think adding student an integrated data system. With a JPA approach, financial aid data to the system also would have the entities that know their data best (the data considerable value, but CSAC is in the midst of contributors) have a direct role in developing and replacing its existing data system. Waiting until www.lao.ca.gov 7 analysis full gutter 2019-20 BUDGET CSAC’s IT upgrade has been completed could help expansion, however, is likely to come with new make its inclusion in the integrated system more challenges and risks. successful. After completing and testing the initial Specify Expectation for JPA to Create Data phases of the integrated system, the Legislature Warehouse and Adopt Certain Policies. In could consider adding data from other agencies, addition, we recommend the legislation specify that such as health and human services agencies. the JPA develop a centralized data warehouse, update the SB 1298 technical work group’s plan Recommendations based on this expectation, and follow CDT’s project In this section, we lay out a recommended series approval process (including for the university of steps for the Legislature to authorize and fund components of the project). the development of an integrated education data Approve One-Time Funding, With Reporting system. Requirements. We recommend the Legislature Recommend Enacting Legislation Creating a approve $10 million in initial funding for the JPA and JPA. Similar to SB 885 (Simitian), we recommend provide associated expenditure authority through the Legislature authorize the creation of a JPA to June 30, 2023. To ensure the Legislature is kept develop and maintain an integrated data system. apprised of the project’s progress, we recommend We recommend the legislation specify that each the Legislature appropriate the funding in the education segment contributing data be a member annual budget act and require the administration but state-level, non-segmental members form a use the section letter process to notify the Joint majority on the JPA. We recommend specifying Legislative Budget Committee prior to allocating that the non-segmental members include any round of funding. We recommend making representatives from the administration as well as approval of initial rounds of funding conditioned on bipartisan representation from the Legislature. the JPA meeting certain milestones, including: Direct JPA to Focus Initially on Data From • Submitting a multi-year work plan to the Core Education Segments and EDD. We Legislature by March 1, 2020. recommend the legislation also define the initial • Submitting a budget to the Legislature by scope and long-term vision of the project. We May 1, 2020. recommend the initial goals be to connect data • Progressing in a timely manner through CDT’s from CDE (K-12), the three public higher education planning phases. segments, and EDD. The legislation could state the longer-term intent that, once those connections • Developing protocols for maintaining data are successfully made, adding data from other privacy and security. entities (such as from early education providers, • Developing protocols for making the data CSAC, and health and human services agencies) available for research purposes. be considered. Eventually, the state might consider adding even more data sources, such as from corrections and private schools. Each phase of 8 LEGISLATIVE ANALYST’S OFFICE analysis full gutter 2019-20 BUDGET CONCLUSION The administration’s goal to create an integrated practices. Our review of the Governor’s specific education data system for California is in line with a proposal to create such a system has identified longstanding priority of the Legislature. Such a data some promising components, but also a number system could give the Legislature a more holistic of shortcomings. Compared with the Governor’s view of the state’s education system and allow proposal, our alternative set of recommendations policymakers to make more informed budget and leverages planning that has already been done, policy decisions. An integrated data system also uses available funding more effectively, reduces could provide more information to educators about the risk of the project failing by requiring it to go what happens to their students after leaving their through established state review processes, and particular education segment, thereby providing strengthens legislative oversight throughout the greater insight into the effectiveness of current development and implementation process. www.lao.ca.gov 9 analysis full gutter 2019-20 BUDGET LAO PUBLICATIONS This report was prepared by Paul Steenhausen and reviewed by Jennifer Kuhn Pacella. The Legislative Analyst’s Office (LAO) is a nonpartisan office that provides fiscal and policy information and advice to the Legislature. To request publications call (916) 445-4656. This report and others, as well as an e-mail subscription service, are available on the LAO’s website at www.lao.ca.gov. The LAO is located at 925 L Street, Suite 1000, Sacramento, CA 95814. 10 LEGISLATIVE ANALYST’S OFFICE