Introduction

Recent comparative research has confirmed that there remain wide variations in the extent of cross-sectional child poverty across countries at otherwise similar stages of development. In this chapter we present the latest evidence on variations in child poverty across the industrialised world, and assess the contributions of family structure, state transfers and market incomes to this variation.

Our results are based on data from the Luxembourg Income Study (LIS) which, at the time of writing, covers some 25 industrialised countries, many with information for several years. We utilise almost all these data in the results presented here. The countries examined include most of the OECD, several of the important non-OECD economies of Eastern Europe (including Russia) and one representative of the newly industrialising countries of East Asia (Taiwan).

The remainder of the chapter is structured as follows. In the next section, we briefly describe the methods we use to measure poverty. Then we show both the ordering of countries by child poverty in the latest available wave of LIS, as well as the poverty trends that can be estimated using all of the LIS data points for each of the countries. We discuss the role of the public sector in affecting poverty rates from a few different points of view and present in the final section some concluding remarks.

The measurement of child poverty

The sharing unit and equivalence scale

Two major decisions that must be made in any poverty study concern the choice of sharing unit (eg, within nuclear families or within households) and the equivalence scale (the needs of different types of sharing units). There is a very large literature that addresses these issues (see for example, Jenkins and Lambert, 1993; Gottschalk and Smeeding, 1997; Jäntti and Danziger, 1999). Our choices on these matters are fairly standard.

Our measure of resources is annual3 disposable income. This includes market incomes and government cash transfers, and deducts income taxes and compulsory social insurance contributions. While this is not a comprehensive indicator of the resources available to the families of children (eg it excludes non-cash services) it remains the best available indicator of cross-national variations in living standards.