Poverty Comparisons with Absolute Poverty Lines Estimated from Survey Data

The objective of measuring poverty is usually to make comparisons over time or between two more groups. Common statistical inference methods are used determine whether an apparent difference in measured statistically significant. Studies relative have long recognized that when the line calculated from sample survey data, both variance and welfare metric contribute estimate. In contrast, studies using absolute lines ignored variance, even estimated data. Including could either reduce increase precision estimates, depending on specific characteristics This paper presents a general procedure for estimating standard error measures Based bootstrap methods, approach can be wide range lines. method applied recent household data Mozambique. When sampling taken into account, errors Foster-Greer- Thorbecke Watts by 15 30 percent at national level, with considerable variability lower levels aggregation. -- Authors' Abstract