State Household Surveys Alive and Well!

February 9, 2009. Pennsylvania was the most recent state to release results from its own state household survey. Despite valuable information on health insurance coverage that is released every year as part of the Census Bureau’s Current Population Survey (CPS), many states still find it essential to administer their own household survey to get more detailed information on the characteristics of the uninsured and more timely information to inform state health policy.

Results from the 2008 Pennsylvania Health Insurance Survey indicate that over 1,000,000 Pennsylvanians lacked health insurance coverage in 2008 and that uninsurance rates across all ages were up slightly from 7.5% in 2004 to 8.2% in 2008.

Similar to national trends, the survey shows an increase in the number of working-aged adults (age 19-64) with 11.6% uninsured in 2008, up from 11% in 2004. Almost three-quarters of the uninsured in Pennsylvania were working-aged adults. Among children (age 0-18), 4.6% were uninsured in 2008, also up slightly from 4.0% in 2004. The recent passage of the SCHIP expansion should provide opportunity to improve kids’ coverage, which may be especially important during this economic downturn.

State surveys allow the flexibility to add state-specific questions to target outreach and enrollment activities. Pennsylvania added interesting questions about knowledge of state public programs and barriers to enrollment. Interestingly, the state estimates that two-thirds of uninsured adults (approximately 580,000) were eligible for coverage through Medicaid or Pennsylvania’s adultBasic program but most did not have any or had little knowledge of the state’s subsidized programs. In addition, demand for the state programs outstrips the supply - there is currently a waiting list for the adultBasic program.

The Pennsylvania survey, conducted from September 2007 through May 2008 by Market Decisions for the Pennsylvania Department of Insurance, contacted 20,222 households and interviewed 49,345 residents in 2008, with a response rate of 38%. It should be noted that response rates are down for all types of telephone surveys. SHADAC’s Mike Davern has studied survey response rates and the methods survey researchers use to get non-responders to respond at a later point in time. He found differences in the demographics of those who responded on the first call and those who responded later. In fact, low response rate might not affect the bias in the estimates you are interested in. See SHADAC’s issue brief, "Are lower response rates hazardous to your health survey?"

More information about Pennsylvania’s survey and the methodology.


Lynn Blewett, Ph.D.