If a researcher wants to show the importance of the problem, citing an estimate that shows the magnitude within a wide band of accuracy may serve the purpose. However, if the estimates will be used to determine the resources needed to address the problem more detailed information about the distribution of the uninsured, their income, age, and use of health resources is essential. The different series of data cited above are therefore not redundant—rather they serve different analytical purposes.
For example, a news account of problems related to people who do not have health insurance might quote an estimate of the size of the uncovered population. Because the purpose is to show the problem broadly, an estimate expressed in the millions—that cites a range with uncertainty of several million—might suffice.
Budgeting for Coverage
However, when planning to address the problem, federal budget officials may need a more precise figure so that they can estimate the cost of extending health insurance coverage—and they may need to know the kinds of conditions likely to be covered, the ages, incomes, employment status and assets of the uninsured.
“Quality” Differs Depending on Use Thus, a survey that approximately reports on the number of uninsured may serve the needs of the news account while not really providing the information that the health insurance planners or budget officials would need.