|
Abstract:
|
Statistical bureaus in many countries conduct periodic sample surveys to obtain socioeconomic data for use in developing indices of interest to policy-makers and public agencies. The fixed periodicity of these surveys is a relic of the days when the primary data collection method was a census. Surveys, we argue, should be done continuously (i.e., as often as is practical and cost-effective) with much smaller samples. The policymakers and agencies are interested in changes in indices, and when continuous sampling uncovers such changes, larger surveys can be done to confirm them with modified scope and allocation. This dynamic approach, in contrast to the static approach reminiscent of a census, provides challenges to survey practitioners but would be more useful to policymakers and other users of data. To illustrate the challenges and opportunities, we consider several examples.
|
