Most household survey data come from a single cross-section of households surveyed at a single point in time. This is useful if the purpose is to get a snapshot of income or poverty, and it does allow for a detailed analysis, for instance, of the proximate determinants of health or malnutrition or income. However, it is rarely possible to get an adequate appreciation of dynamic effects how incomes in a household rise and fall, how households form and re-form without panel data.