During the 1990s, the Intercensal Estimates Program of the US Census Bureau prepared and disseminated national, state, and county estimates of the Hispanic Population. These estimates were developed independently of the Current Population Survey, using component and ratio techniques. Evaluations of these data compared to the Census 2000 show deficiencies in both the overall national estimate as well as the estimates of the geographic distribution of the Hispanic Population.
This paper will examine the methods used to estimate the Hispanic population during the 1990s and present a brief evaluation of how these estimates compared to the Census 2000 results. The paper will discuss the challenges involved in preparing these independent estimates, including the difficulties encountered in using the traditional sources of demographic accounting data.
To address these difficulties, various approaches are being developed, including the use of modeling and surname algorithms. The paper also examines the possible use of survey and other administrative data to complement the traditional demographic accounting sources.
|