Tuesday, August 29, 2006
Snapshots vs. movies of health insurance coverage
Here's the problem with this type of information from the Census -- to contribute to Eric Black's big question today -- it's a snapshot, taking stock of who has how much money, who has health insurance coverage, etc., at a point in time. The stock of people without health insurance at one time might be larger or smaller than the stock at a previous period in time, but that doesn't tell us who those people are. What we want to know is not just how many people are without insurance at a point in time, but how long it lasts.
Moreover, as this study from the Census Bureau makes clear, the data used to make these claims about rising lack of health insrance coverage is from the Current Population Survey that is less accurate than the Survey of Income and Program Participation. Only the latter gets at the dynamics by asking the same individuals about health insurance over time.
The difference between the CPS studies and the SIPP studies is like the difference between a photograph and a movie. Which one gives you a better representation of health insurance coverage?
Moreover, as this study from the Census Bureau makes clear, the data used to make these claims about rising lack of health insrance coverage is from the Current Population Survey that is less accurate than the Survey of Income and Program Participation. Only the latter gets at the dynamics by asking the same individuals about health insurance over time.
Although both surveys are household surveys conducted by the United States Census Bureau, they are designed to meet different needs and, hence, have different sample sizes, interview techniques, sample compositions, and survey reference periods. Accordingly, the two surveys produce varying health insurance coverage rates.As best I can tell, the 2004 SIPP data for health insurance coverage is not yet available. The 1996-99 wave of participants showed 8.3% of individuals without insurance for an entire calendar year, while another 13.3% had no insurance for at least one month in that year.
The CPS ASEC, which collects annual information, found that 83.6 percent of people were covered by health insurance for some or all of 1998. The SIPP, which collects monthly information, found 92.0 percent of people covered by health insurance for at least 1 month of 1998.2 Since the SIPP collects monthly information and allows us to see changes from month to month, SIPP may be closer to the truth.
The difference between the CPS studies and the SIPP studies is like the difference between a photograph and a movie. Which one gives you a better representation of health insurance coverage?
