Authors: Leo Azzollini, Richard Breen, and Brian Nolan
Institution: Stone Center Working Paper Series. no. 80
Date: December 2023
Abstract:
Changes in household structures and employment patterns alter the balance between households with an above- versus a below-average poverty risk while also affecting relative income poverty thresholds. Examining eleven countries for which suitable microdata is available from the Luxembourg Income Study (LIS) back to the mid-1980s shows that patterns of change in household composition and employment exhibited some common features but also very substantial variation. The share of single adult households rose in most countries, couples with no or only one person in paid work fell in most, while couple households with two earners increased in a majority but not in Denmark, Norway and the USA and only modestly in Hungary and the UK. A counterfactual exercise assessed the impact of these changes in composition on relative income poverty rates by reweighting the 2019 samples to impose the composition structure observed in 1986. In the absence of these composition changes the relative poverty rate in 2019 would have been a good deal higher in Germany, Greece, and Italy, and especially in Israel and Spain. Composition changes had only a modest impact in the UK and made very little difference in Denmark, Hungary, and the USA, while working to increase the relative poverty rate in Czechia and Norway. This reflected the varying scale and nature of the composition changes seen across these countries. Their impact included driving up the relative poverty threshold (except in the USA), and if this effect is discounted the composition shift over the period would have had a greater poverty reduction impact in most countries, especially in Israel, Italy and most powerfully in Spain.