Frequently Asked Questions

Do you have questions about IDSs and the AISP Network? We have answers.

Click on a question to reveal its answer.

Integrated Data Systems (IDS) link administrative data across government agencies to improve programs and practices through evidence-based collaboration.  Click here for more information.

Administrative data are records that are maintained by public and private agencies to track the business activities of those agencies.  They can include vital records, like birth and death certificates, medical records, school district records, earnings and employment data, housing assistance, justice programs, etc., as well as privately collected records such as credit scores.

People’s day-to-day experiences and interactions don’t exist in siloed spaces. However, the data collected by different government agencies and service providers are typically not connected. This means that it’s difficult to understand how policy or program changes in one agency affect a group’s outcomes in another agency. Integrating administrative data across government agencies changes this. For example, because of IDS, we know that permanent supportive housing for chronically homeless individuals reduces their rate of incarceration and hospitalization, while also decreasing government costs and increasing well-being.

IDS allow counties, states, and cities to evaluate the programs and policies that serve their constituents so decision-makers can implement strategies that best address their community’s needs. Furthermore, integrated administrative data can also be used in the development of new, innovative responses to societal problems. Through social policy experimentation, IDS can more quickly determine whether this newly implemented program has the intended effect.

The AISP Network is a professional group of individuals within county organizations, state agencies, and universities that operate integrated data systems (IDS) across the U.S. They are leaders in the field of integrated data systems and have mature, functioning IDS capable of producing actionable intelligence to guide policy and practice decision-making. Our professional network includes the following sites:

  • Center for Innovation through Data Intelligence (CIDI), NYC
  • Chapin Hall at the University of Chicago
  • City of Philadelphia, Office of the Deputy Mayor for Health and Opportunity
  • Cuyahoga County ChildHood Integrated Longitudinal Data (CHILD)
  • Department of Human Services (DHS), Allegheny County, PA
  • Health and Demographics, South Carolina Revenue and Fiscal Affairs Office
  • Institute for Social Capital, Inc. Greater Charlotte Region, North Carolina
  • LA County, CA
  • Milwaukee, WI: Data Share
  • Providence Plan
  • State of Michigan
  • USF: Policy and Services Research Data Center (PSRDC)
  • Washington State

For more information, please visit the Network page.

Your organization can request to become part of the AISP network by following these guidelines.

A) Establish ongoing data sharing and integration from at least three agencies or program providers that include formal documentation of the relationships (e.g., signed Memoranda of Understanding (MOU).

B) Create a formal structure (governance) to review and approve specific projects that utilize the integrated data system at your site.

C) Forming an ongoing infrastructure that exists beyond a single project. This should include analysts and analytic capabilities through partnership and funding for on-going operations.

AISP network membership indicates that a site has the infrastructure to participate in multi-site research projects.  Please refer to the AISP Network Membership page for more information.

By comparing data across various agencies, researchers can ask probing questions about issues that have stymied public agencies’ past efforts to improve policies. Integrated data reveal patterns of risk and resilience, which, in turn, more readily links cause and effect, thereby allowing officials to test more targeted interventions and higher-impact policies. IDSs also test social policy innovations through high speed, low-cost randomized control trials (RCTs) and other quasi-experimental approaches (e.g., observational studies, benefit cost analysis, case control studies). IDS can also be used to form a multi-system continuous quality improvement approach to program performance assessment.

People’s day-to-day experiences and interactions don’t exist in siloed spaces. However, the data collected by different government agencies and service providers are typically not connected. This means that it’s difficult to understand how policy or program changes in one agency affect a group’s outcomes in another agency. Integrating administrative data across government agencies changes this. For example, because of IDS, we know that permanent supportive housing for chronically homeless individuals reduces their rate of incarceration and hospitalization, while also decreasing government costs and increasing well-being.

IDS allow counties, states, and cities to evaluate the programs and policies that serve their constituents so decision-makers can implement strategies that best address their community’s needs. Furthermore, integrated administrative data can also be used in the development of new, innovative responses to societal problems. Through social policy experimentation, IDS can more quickly determine whether this newly implemented program has the intended effect.

You can build a successful IDS within your organization by fully addressing the below components.

  1. Governance Process: Every integrated data system requires a formal governing board, composed of representatives from the “source” agencies who own or control the various databases to be included.  These agencies have legal responsibilities for their data, and must approve individual data requests.  The governing board may also include data experts or legal advisors.
  2. Data Integration: The critical methodology for creating integrated data systems is the process of record linkage, which refers to the joining or merging of data on the basis of common data fields. The data fields are usually personal identifiers (e.g., name, birth date, social security number, or encrypted versions of those identifiers).
  3. Data Quality and Scientific Issues: Accurate data are essential to maintaining the integrity of research. In order for an integrated data system to produce the most dependable results, the governing board overseeing the IDS must establish appropriate methods for assessing the reliability and validity of data elements to maximize the utility of the information they contain. Research questions must also be able to be tested and answered using appropriate methods for integrated data. Scientific expertise frequently aids in the interpretation of results and plays a part in the translation of findings into policy and practice change.
  4. Partnerships with Stakeholders:  Integrated data systems exist in order for providers, researchers, and agency leaders to come together with the goal of answering complex questions about how health, education, workforce and human services can effectively and efficiently better meet the needs of the public. Because of this, integrating data is a process of evaluation, discovery, and improvement by stakeholders with various perspectives. The essence of an IDS lies in partnerships across different stakeholder groups that seek to set priorities, lead inquiry, and translate results into actionable intelligence that betters policy and practice. Partnerships with foundations and other funders can also provide funding for the ongoing operations of an IDS.
  5. Research Administration: Establish successful administrative processes for sharing data.

Laws and regulations at all levels of government regulate the confidentiality and privacy of information collected and stored by government agencies. Federal regulations are the most explicit in providing detailed information about what can and cannot be done with protected health, social service and education data, for example, whereas other social policy areas may be further regulated at the state or local level (e.g., child welfare, juvenile justice, or homelessness). The federal Privacy Act of 1974, 5 U.S.C. § 552a (2000), is the omnibus “code of fair information practices” that regulates the collection, maintenance, use, and dissemination of personal information. The Privacy Act is designed to balance the government’s need to maintain information about individuals with the rights of individuals to be protected against unwarranted disclosure of personal information (i.e., any data element that can be used to identify the individual like names, Social Security numbers, and addresses). FERPA and HIPAA are the federal laws governing education and health information, respectively.  Each of these federal laws include exemptions or exceptions that permit public agencies to share data for purposes of research, audit and evaluation, provided that appropriate data security policies and procedures are in place, and that identifying data are either not shared (de-identified research files are created) or if shared they are not redisclosed (governed by a business agent agreement or other contract).

Numerous laws and regulations provide broad protections against the use of private information and determine how, when, and for what purpose these data can be integrated and shared between public agencies. Building a system that relies on IDS requires that these legal issues are attended to through the creation of a written memorandum of understanding (MOU) between IDS partners. These MOU’s provide the collaborative foundation of the IDS and accomplish two important objectives. First, they make it the top priority to protect the private information of individuals being served by the respective service agencies. Second, they respect the rights and responsibilities of the agencies that collect the private information to provide services and to use these data to learn how they can improve the quality of the services they provide.

A Data Use Agreement (DUA) is a legal binding agreement between a department and an external entity (e.g., contractor, private industry, academic institution, other Federal government agency, or state agency), when an external entity requests the use of personal identifiable data that is covered by a legal authority.  A DUA is required when identified or potentially identifiable data are shared by an IDS with a researcher or other agency.

Actually, it’s much less costly than traditional research methods. Since administrative data are already routinely collected by service-providing agencies, data collection costs are virtually non-existent. For example, a study following 10,000 over 10 years could be done in 18 months for $150,000, whereas a similar, traditional project would likely cost millions of dollars, and take 15 years.

The most common way to fund the development of an IDS is through a grant (generally from a local foundation or government agency) to perform a demonstration project that addresses a pressing policy issue in your community. This provides start-up funding for the operations, infrastructure, data integration, and analysis. Demonstration projects illustrate the value of IDS to local stakeholders, and should lead to future project requests and funding. It’s important to build your community’s IDS to address the priority needs of the community and participating agencies. While it’s easy to get lured into buying an expensive, state of the art technology system, it’s more important to have a system that meets your organization’s needs and allows for future growth.

Some of the reasons that IDS positively impact the community are:

  1. Allowing government agencies to integrate various databases and bridge the information and knowledge gaps that have traditionally formed between them.
  2. Providing researchers the opportunity to compare data across various agencies and ask probing questions about issues that have stymied public agencies’ past efforts to improve policies.
  3. Revealing patterns of risk and resilience, which, in turn, more readily links cause and effect, allowing executive leaders, policy makers, and policy analysts to test more targeted interventions and higher-impact policies.
  4. Developing and testing social policy innovations, such as alternatives to juvenile incarceration, programs to reduce truancy or improve educational outcomes and career readiness, etc.

Yes, we can provide consultation on the following topics:

  • Organization’s basic readiness (e.g. required partnerships, the entity that will host the IDS, funding, research priorities, and systems development) the governance process, and legal issues.
  • We also coordinate consultation with network members who have extensive IDS experience on a wide variety of topics.
  • You can go to the About Us/Services page to complete our IDS Assessment Tool

The essential components of an IDS are securing and maintaining legal agreements, implementing data analytics management and processes, established governance processes, and developing political and economic elements to sustain the IDS. For more information, please go to IDS/Establishing an IDS page.

Yes, IDSs test program interventions by:

  1. Assessing their impact on utilization uptake and outcomes
  2. Creating comparison groups of people not exposed to an intervention to estimate the potential effects of the intervention
  3. Monitoring individual-level impacts of population-level interventions (e.g., public service announcements, educational campaigns)

The common analytic uses of IDS can be used to identify where an intervention will be most effective from observational studies, patterns of service use by sub-populations, tracking outcomes associated with interventions or intervention durations, identifying risk or protective factors in data sources that may mediate or impact an intervention, looking at how interventions in a single agency impact outcomes across multiple agencies, identifying high-risk groups or multi-system service users, including by neighborhood or other geographic region.

Yes, there are plenty of examples on our website that showcase how IDS have changed polices or improved outcomes for people.  Please go to the Resources/Examples page or the IDS Uses page.