Data Focus
The U.S. Census Bureau securely provisions more than 300 restricted-use data assets for evidence-building research conducted by qualified researchers on approved projects through the Federal Statistical Research Data Centers (FSRDCs). This section highlights how individual restricted-use data assets are used in the FSRDC network and showcases examples of the scientific and programmatic benefits generated through secure access to confidential microdata.
New examples will be added periodically to highlight the expanding impact of FSRDC-enabled research.
FSRDC Research Using Data from the Medical Expenditure Panel Survey—Insurance Component (MEPS-IC)
The following example illustrates how researchers use restricted-use data in the secure FSRDC environment to generate both scientific insights and programmatic benefits, including operational improvements to surveys, data quality, and statistical methods.
About This Data Asset Example
This profile highlights how one restricted-use data asset supports evidence-building research through the FSRDC network. Additional data asset examples will be added over time to demonstrate the breadth of datas available to qualified researchers.
Data Overview
The Medical Expenditure Panel Survey – Insurance Component (MEPS-IC) is an annual survey conducted by the U.S. Census Bureau for the Agency for Healthcare Research and Quality (AHRQ), part of the U.S. Department of Health and Human Services. The survey is the nation’s most comprehensive source of data on employer-sponsored health insurance, producing estimates at the national, state, and metropolitan statistical area levels.
Since 1996, the MEPS-IC has surveyed a large rotating sample of approximately 40,000 private-sector establishments and 3,000 state and local governments annually. The survey collects detailed information on employer-sponsored health insurance offerings, including plan types, benefits, premiums, employer and employee contributions, eligibility requirements, enrollment levels, and employer characteristics.
How Researchers Use MEPS-IC Data in the FSRDCs
Policymakers, researchers, and industry analysts commonly use tabular data from the MEPS-IC published online by AHRQ to track trends in health insurance coverage, costs, and plan design. In addition, the Bureau of Economic Analysis uses MEPS-IC microdata to estimate aggregate employer contributions to group health insurance, a key component of labor income in U.S. Gross Domestic Product.
Because of confidentiality protections, establishment-level MEPS-IC microdata are not publicly released. However, approved researchers may securely access restricted-use MEPS-IC data through the FSRDCs. Within the secure FSRDC environment, researchers can link MEPS-IC data to other confidential Census Bureau data assets, such as the Longitudinal Business Database (LBD), which contains additional firm- and establishment-level information including employment, payroll, and revenue over time. These linkages enable analyses that would not be possible using publicly available data alone.
Research Benefits to the Census Bureau
Researchers granted access to Census Bureau restricted-use data for statistical purposes, including MEPS-IC microdata, are legally required to provide research benefits to the Census Bureau. These benefits generally fall into two categories: programmatic benefits, which improve survey design, data quality, and statistical operations, and scientific benefits, which generate new population estimates and policy-relevant insights.
Research using MEPS-IC microdata has produced substantial programmatic benefits for the Census Bureau, particularly in improving data quality, survey design, and methodology.
Examples include:
Improving data quality and consistency
Researchers identified inconsistencies in reported employer health insurance plan provisions, including discrepancies across establishments within the same firm, and proposed corrections to enhance data accuracy.
Cross-validating data with linked sources
Studies validated MEPS-IC data using linked Census Bureau data assets such as the LBD, as well as external data sources including the Kaiser Family Foundation/Health Research & Educational Trust Employer Health Benefits Survey.
Reducing respondent burden while enhancing survey content
Cross-validation of firm age information between the MEPS-IC and the LBD supported removal of a redundant survey question. This change created space for new survey content while maintaining or improving data quality without increasing respondent burden.
Evaluating survey design changes
Researchers assessed how questionnaire modifications affected item nonresponse rates and survey representativeness, informing improvements to survey operations.
Enhancing contextual information
Researchers linked MEPS-IC data to aggregated external sources such as unemployment rates, County Business Patterns, American Hospital Association surveys, and Medicaid policy data to provide additional market, demographic, and regulatory context.
Advancing statistical processing methods
Research led to new imputation methods for addressing item nonresponse, analysis of nonresponse patterns by firm size, industry, and geography, and identification of firm characteristics associated with nonresponse to improve survey targeting.
MEPS-IC research conducted through the FSRDC program has generated new empirical findings and policy-relevant analyses. Selected examples include:
Abraham, J.M., A.B. Royalty, and C. Drake (2019). The impact of Medicaid expansion on employer provision of health insurance. International Journal of Health Economics and Management, 19:317–340.
Using 2010–2015 MEPS-IC microdata augmented with Local Area Unemployment Statistics and Medicaid eligibility data, researchers found that Medicaid expansion under the Affordable Care Act did not produce large-scale changes in employer-sponsored insurance offerings, eligibility, employee acceptance rates, or employee premium contributions.
Karaca-Mandic, P., R. Feldman, and P. Graven (2018). The role of agents and brokers in the market for health insurance. Journal of Risk and Insurance, 85(1):7–34.
Using 2008 MEPS-IC data and county-level measures of insurance agent and broker competition, researchers found that increased competition was associated with higher rates of employer insurance provision and lower premiums, suggesting that competition helps employers navigate complex insurance markets more efficiently.
Lin, K.-H., S. Bondurant, and A. Messamore (2018). Union, premium cost, and the provision of employment-based health insurance. Socius, 4:1–11.
Using 1999–2012 MEPS-IC microdata linked to the LBD and other data assets, researchers found that employers were more likely to offer health insurance when workers were unionized and when establishments operated in states with higher unionization rates and union-friendly laws.
Abraham, J.M., R. Feldman, and P. Graven (2016). Employers’ changing economic incentives to offer health insurance under the Affordable Care Act. American Journal of Health Economics, 2(3):273–299.
Using 2008–2010 MEPS-IC microdata augmented with external information, researchers found that small employers were the most responsive to changes in the tax treatment of health insurance premiums, while larger firms showed less sensitivity.
Acknowledgment: The Census Bureau thanks Joseph Ballegeer and Kenneth Zahringer for their review of scientific benefits produced using MEPS-IC microdata.
Interested in doing research at an FSRDC?
Contact the FSRDC administrator nearest you or reach out to the Census Bureau at [email protected].
