In the last few years several surveys have been conducted which have had as one on their goals to tell us what happens to households and families over time. These include the National Medical Care Expenditure Survey (NMCES), the National Medical Care Utilization and Expenditure Survey (NMCUES) and the Survey of Income and Program Participation (SIPP). The process of obtaining longitudinal estimates for such units presents some important questions that either are not encountered or are easier to answer for cross-sectional estimation or longitudinal person estimation. The following are three key questions:

1. Since the composition of households and families can and does change over time, which changes should allow the unit to be considered still continuing and which mark the dissolution of the unit?
2. What operational rules should be used to determine which households, families and individuals are to be followed over time, and what retrospective questions should be asked of individuals who join the sample after the beginning of the panel?
3. What weighting procedures should be employed to obtain weights that yield unbiased estimates, and how should the weights be adjusted to reduce the variances and biases of the estimate?