International analysis of child poverty – ukmod/euromod modelling

2. Data and Methodology

2.1 Choice of countries, data, and time period for analysis

Based on UNICEF and EUROSTAT data, four comparator countries were selected for analysis: Croatia, Denmark, Finland, and Slovenia. These countries were selected because:

• they currently have child poverty rates that are close to the Scottish Government targets (Denmark, Finland, Slovenia) and/or have seen significant reductions in child poverty rates in recent years (Croatia, Slovenia);
• they possess similar features to the Scottish context, such as similar demographic size and profile, either now or in the past;
• they are part of the European Union (permitting use of standardised data and simulation methods).

This set of countries considered here is also considered by the parallel review of qualitative evidence.

The current study addresses the research goals defined above by decomposing differences in child poverty rates between Scotland and the comparison countries into three components: demographics, labour markets, and tax-benefit systems. The analysis compares Scotland and the four selected countries with reference to the estimated child poverty rates in 2024. Additionally, and only for Slovenia and Croatia, the analysis explores changes in child poverty rates between 2011 and 2024.

The analysis is conducted using UKMOD^[7]/EUROMOD^[8] data. Use of these data is advantageous for a cross-country analysis as considered here, as the country-specific data have been subject to pre-processing to facilitate comparisons between them.

UKMOD data for Scotland are derived from the Family Resources Survey (FRS), which reports incomes observed during the same (fiscal) year of the survey. Processed data for UKMOD are available at annual intervals from 2012 to 2022 inclusive.^[9]

EUROMOD data for all four comparator countries are derived from the EU Survey of Income and Living Conditions (EU-SILC). EU-SILC reports incomes observed during the (calendar) year preceding the survey. The latest available input data for EUROMOD are for 2022, which reports incomes observed in 2021. As these data were potentially affected by the Covid-19 pandemic, we use EU-SILC data reported for 2019 (2018 incomes).^[10] These technical details are summarised in the Appendix.

2.2 Analytical methodology

We decompose international differences in child poverty rates between Scotland and comparison countries into three components:

1. Demography: age, gender and household composition of the respective populations;

2. Labour markets: employment rates, working hours and wages;

3. Tax-benefit systems: direct tax and benefit transfer payments.

The analysis focuses on relative poverty rates among dependent children. As per the Child Poverty (Scotland) Act 2017, relative child poverty is defined as the fraction of children living in a household with equivalised income less than 60% of the respective population median. Incomes are equivalised using the revised OECD equivalence scale, in line with standard practice.

The definitions of poverty considered for the current analysis do, however, depart from official statistics in a number of ways:

• Official statistics focus on poverty rates measured with respect to disposable incomes. The current analysis is based on measures of both original and disposable incomes, which are contrasted to evaluate the influence of public transfers on poverty.
• Before Housing Costs (BHC) income is considered for analysis, rather than After Housing Costs (AHC) income as is commonly considered in official distributional statistics for the UK to ensure consistency between countries.
• Official poverty lines in Scotland are defined with respect to UK median income. The current analysis, in contrast, evaluates poverty lines based on the distribution of income in Scotland alone, in line with the treatment of comparator countries.

Disparities with officially reported statistics may also arise where the current analysis focusses on modelled income rather than measures of income reported in publicly available microdata or official data sources.

This analysis focuses primarily on differences between measures of poverty, rather than magnitudes of individual poverty measures. Hence, what is most important here is that the measures of poverty that are compared are consistent across countries.

Further detail can be found in the Appendix A.