Policy Options to Encourage All-Payer Claims Databases

After a 2016 Supreme Court decision, policymakers must re-evaluate strategies for collecting health care data.

A close-up of a stethoscope. (Getty/Universal Images Group)
A close-up of a stethoscope. (Getty/Universal Images Group)

A little-known 2016 U.S. Supreme Court decision, Gobeille v. Liberty Mutual Insurance Co.,1 dealt a blow to state efforts to lower health care costs. In this case, the court struck down a Vermont law that required all employers to submit data to the state’s all-payer claims database (APCD), finding that an existing federal law pre-empted the state statute.2 Unlike prior Supreme Court decisions that upheld key provisions of the Affordable Care Act (ACA), Gobeille went barely noticed. The decision, however, runs contrary to a key goal of the ACA: lowering health care costs and improving the quality of care.

This issue brief outlines both short-term state-based actions and long-term federal policy options to mitigate and ultimately undo the limits that the Gobeille decision placed on states.

APCDs encourage high-quality, lower-cost health care

The ACA included a variety of new policies and investments designed to encourage the delivery of high-quality, lower-cost care. For example, the Center for Medicare & Medicaid Innovation (CMMI) tests the effects of various payment and delivery system models on Medicare and Medicaid program expenditures, as well as the quality of care under those programs.3 But the ongoing stalemate in Washington, D.C., about the future of the ACA makes it less than certain that policymakers will embrace additional steps to lower health care costs. For this reason, both state-level and private reforms and innovations have become even more essential to achieving the ACA’s goal of lowering costs and improving care across the nation’s fragmented health care system.

As of December 2016, 16 states had functioning APCDs.4 Generally, APCDs contain medical claims, pharmacy claims, and dental claims, as well as additional information about provider and patient demographics from most public payers—including Medicare and Medicaid—and private payers.5 State APCDs come in various forms: state led; a combination of public and private efforts; and those run by private nonprofits.6

These databases can give state policymakers, private payers, and academics critical data that inform decisions about health care cost and quality—including health and payment reform changes, care delivery design, transparency, population health, and policy and budget development.7 The audience for APCDs is not made up of patients themselves, but rather health care payers, academics, and policymakers, who can use these data to evaluate health care cost and quality, ultimately benefiting patients. Analyses of the information included in these systems can inform provider network decisions and payment rates, ideally resulting in policies that drive consumers toward higher-value providers.8 State regulators can also use these data to assist in rate review to make and keep premiums affordable. For example, Colorado used APCD data to assess differences in pricing for common procedures and how utilization of health care services changes over time, while Oregon used such data to help guide its health system transformation efforts in the state, resulting in $139 million in savings from 2013 to 2014.9

The Gobeille decision and new limits on state APCDs

Ideally, all-payer claims databases contain enough data to allow policymakers and private health care payers to consider quality, utilization, and cost trends across the entire health care system. Too often, these groups must use fragmented data sets that might be incomplete and not representative of the broader population. The Gobeille decision only makes this problem worse.

In Gobeille, Liberty Mutual, which self-funds its employee health plan, challenged Vermont’s APCD law that required all health insurers and plan administrators to submit data to the state for inclusion in the APCD.10 Liberty Mutual argued—and the court agreed—that the Employee Retirement Income Security Act of 1974 (ERISA) pre-empted the state requirement for self-funded or self-insured employers.11 Self-funded or self-insured employers function as the insurer and bear the risk of employees’ health care costs instead of purchasing health insurance coverage from insurance companies for their employees.

ERISA established uniform standards for both pension plans and other benefit plans such as health and disability benefits following a number of large pension plan failures.12 ERISA also includes sweeping pre-emption language that severely limits the requirements states may impose on self-funded plans; it pre-empts state laws that relate to any employee benefit plan—including health plans—except for state laws that regulate insurance.13 And under ERISA, employers offering self-insured employee benefit plans cannot be considered insurers.14 The result is that while fully insured employer-sponsored plans are subject to both state insurance law and federal law, self-insured plans are subject only to federal law, with very few exceptions.

After the Gobeille decision, states cannot compel self-funded employers or their third-party administrators to submit data for inclusion in APCDs. In 2017, approximately 151 million people had employer-sponsored insurance,15 and 60 percent of these insured employees were covered by a self-insured health plan.16 Larger firms are more likely to be self-insured, with 91 percent of covered employees in businesses with 5,000 or more employees self-funded.17 Of the next-largest employers—those with 1,000 employees to 4,999 employees—81 percent of covered employees were in self-funded plans.18 By comparison, only 15 percent of employees in businesses with fewer than 200 employees were self-funded.19 Without data from these plans, APCDs will be at best less comprehensive and less useful; at worst, they will be potentially misleading.

In Vermont, for example, a higher percentage of nonelderly women—59 percent—than nonelderly men—55 percent—are covered by employer-sponsored plans.20 If self-funded plans are not required to submit data to the state APCD, this could alter the demographic mix of APCD information, potentially skewing quality and cost results.

Long-term policy recommendations

The easiest, most practical response to the Gobeille decision would be for Congress to pass and the president to sign into law changes to ERISA that would allow states to require self-funded employers or their third-party administrators to submit data for inclusion in state-based all-payer claims databases.

In addition to this targeted fix, federal policymakers should enact legislation to build a national APCD. First, some states may not have the resources to build their own APCDs, and a national system would allow policymakers and researchers in those states to access these data. Second, a national APCD would allow policymakers and researchers to access the most comprehensive set of data possible, allowing for additional study of price and quality variations across areas and different populations that may be spread across different states.

A number of private organizations already house claims data,21 so the secretary of health and human services should certify one or more organizations as qualified to receive federal grants to build the database. Federal law must also condition grants upon showing that the private organization can adequately secure and protect these data. In addition to these startup grants, the organization maintaining the APCD should receive user fees from states, payers, and other researchers who access the data.

Short-term policy recommendations

Given congressional inaction on key issues, any legislative fix is likely years away. Moreover, even if the Department of Labor (DOL) had clear authority to compel self-funded plans to submit claims data, past actions have shown that the Trump administration has zero interest in improving the health care system.22 The DOL issued a proposed rule in 2016 that considered changes that could be necessary after Gobeille, but it did not directly address APCDs and was never finalized.23

In the absence of federal action, third-party administrators could also voluntarily submit data to entities, such as the Health Care Cost Institute, that actively maintain databases of information. Several insurers already do this,24 with Blue Cross Blue Shield being one of the notable absences. States should also create incentives to encourage self-funded employers and their third-party administrators to submit claims data. One obstacle to data submission is administrative burden; therefore, uniform data submission standards across states with APCDs could encourage participation. Today, a number of states, vendors, and payers are working with the APCD Council to create a common data layout.25 Uniform data collection would make it much less administratively burdensome, and therefore potentially more enticing, for covered entities to submit comprehensive claims data. States could work with willing plans to gather the most comprehensive data possible.

Second, states or the federal government could also help self-funded plans and their third-party administrators offset the costs of submitting data or offer them a reduced fee when they try to access data from an APCD. While this requires an upfront investment, gains made through collection and analysis of these data should lower the cost and improve the quality of health care over time, saving both federal and state funds. Potential sources of federal funding could be the State Innovation Models (SIM) initiative or CMMI. State-generated funding could come from new or existing public health efforts such as soda or cigarette taxes.

To offset these costs, states could consider seeking funding from the Centers for Medicare & Medicaid Services (CMS) under a Delivery System Reform Incentive Payment (DSRIP) program. DSRIP programs are a type of Section 1115 waiver—named for part of the federal Medicaid law—that allow states to adopt Medicaid policies that differ from usual federal Medicaid requirements.26 Originally, DSRIP initiatives were primarily used for funding for safety net hospitals, but states are now using them to test a range of payment and delivery system reforms.27 Approved DSRIP programs largely focus on four areas, starting with infrastructure development and system redesigns, followed by clinical outcomes improvements, and ending with population-focused improvements.28 For example, CMS-approved DSRIP programs have provided funding for investments in statewide data analytics and infrastructure improvements.29 APCDs are precisely the type of infrastructure investments that states can use to improve clinical outcomes and population health.

Alternatively, if states decide to take a more limited approach to data collection post-Gobeille, they could identify a limited number of high-cost or high-volume items and services in both their Medicaid programs and employee and retiree health plans. States then could require health care providers, such as hospitals, to submit data from all payers on the selected items and services. This solution would be in line with the Gobeille decision, as it requires providers, not employers or third-party administrators, to submit data. This approach cannot substitute for an APCD, but it does allow states to continue to assess utilization, cost, and quality in parts of their health care systems.


A key challenge for health care policymakers to tackle in the coming years is how to continue to lower costs and improve quality in the health care delivery system. The ACA took a number of important steps toward this goal, but since the 2016 election, no additional progress has been made. States must continue to think of innovative ways to work around the federal stalemate, including steps to keep APCDs as comprehensive as possible. Ultimately, however, federal policymakers will need to step in to close existing gaps.

Maura Calsyn is the managing director of Health Policy at the Center for American Progress.

This publication was made possible in part by a grant from the Peter G. Peterson Foundation. The statements made and the views expressed are solely the responsibility of the Center for American Progress.


  1. Gobeille v. Liberty Mutual Insurance Co., 136 S. Ct. 936 (2016).
  2. Ibid.
  3. Centers for Medicare & Medicaid Services, “The CMS Innovation Center,” available at (last accessed February 2018).
  4. National Conference of State Legislatures, “Collecting Health Data: All-Payer Claims Databases,” available at (last accessed April 2018).
  5. Ibid.
  6. The Office of the National Coordinator for Health Information Technology, Claims and Clinical Data Integration: All Payer Claims Data (U.S. Department of Health and Human Services, 2016), available at
  7. Ibid.
  8. Lydia Mitts, “Price Transparency in Health Care: An Introduction” (Washington: Families USA, 2014), available at
  9. The Office of the National Coordinator for Health Information Technology, Claims and Clinical Data Integration: All Payer Claims Data.
  10. Gobeille v. Liberty Mutual Insurance Co., 136 S. Ct. 936.
  11. Ibid.
  12. U.S. Department of Labor, “ERISA at 40 – Four Decades of Protecting America’s Employee Benefits,” available at (last accessed April 2018).
  13. 29 U.S.C. § 1144(a)-(b)(2).
  14. Metropolitan Life Ins. Co. v. Massachusetts, 471 U.S. 724 (1985).
  15. Kaiser Family Foundation and Health Research & Educational Trust, “Employer Health Benefits 2017 Annual Survey” (2017), available at
  16. Ibid.
  17. Ibid.
  18. Ibid.
  19. Ibid.
  20. Kaiser Family Foundation, “Vermont: Health Coverage & Uninsured,” available at (last accessed March 2018).
  21. See, for example, Health Care Cost Institute, “2016 Health Care Cost and Utilization Report” (2018), available at
  22. Sam Berger, “3 Ways that States Can Stop Ongoing Health Care Sabotage,” Center for American Progress, January 9, 2018, available at
  23. Employee Benefits Security Administration, Internal Revenue Service, and Pension Benefit Guaranty Corporation, “Proposed Rules,” Federal Register 81 (140) (2016): 47533–47681, available at
  24. Health Care Cost Institute, “Data Contributors,” available at (last accessed April 2018).
  25. All-Payer Claims Database Council, “Standards,” available at (last accessed April 2018).
  26. MaryBeth Musumeci and others, “Section 1115 Medicaid Demonstration Waivers: The Current Landscape of Approved and Pending Waivers” (Menlo Park, CA: Kaiser Family Foundation, 2018), available at
  27. Alexandra Gates and Robin Rudowitz, “An Overview of Delivery System Reform Incentive Payment (DSRIP) Waivers” (Menlo Park, CA: Kaiser Family Foundation, 2014), available at (last accessed April 2018).
  28. Ibid.
  29. Center for Health Care Strategies, Inc., “Delivery System Reform Incentive Payment (DSRIP): State Program Tracking” (2016), available at

The positions of American Progress, and our policy experts, are independent, and the findings and conclusions presented are those of American Progress alone. A full list of supporters is available here. American Progress would like to acknowledge the many generous supporters who make our work possible.


 (Maura Calsyn)

Maura Calsyn

Former Vice President and Coordinator, Health Policy