Implementing the First ISTAT Survey of Homeless Population by Indirect Sampling and Weight Sharing Method

Abstract

The Italian National Institute of Statistics carried out the first survey on the homeless population. The survey aims at estimating the unknown size and some demographic and social characteristics of this population. The sample strategy for the homeless survey refers to the theory of indirect sampling, based on the use of a sampling frame indirectly related to the target population. In fact, the methodological strategy used to investigate the homeless population could not follow the standard approaches of the official statistics usually based on the use of population lists. In the general context of hard-to-reach population methods, the survey used a sample approach based on a time-location sampling, in which target population units are reached through the random selection of places (the centers) and instants of time. The availability of information about centers providing services to people living in extreme poverty was the fundamental information for the design of the survey. Following the indirect sampling approach, the estimation is performed through the “weight-sharing method,” based on the links connecting the frame of services with the population of the homeless.