Demographic change has motivated various policy interventions to increase different margins of labour supply. This paper evaluates the recent ``flexi-job reform'' from Belgium, which introduced tax-exempt secondary jobs for specific sectors for near-to full-time workers. Using panel data for the years 2012--2019, I analyze the reform's impacts on labour supply, demand and wages for workers and firms using difference-in-differences estimation. The results show a small increase in both secondary working time and secondary wages among full-time workers and a larger increase in working time among the high-response subset of individuals with similar jobs prior to the reform. I also find strong effects on the firm side as firms in affected sectors restructure their workforce towards more secondary work. Taken together, this study shows that there exists a modest potential to increase labour supply on the full-time margin and suggests that hours constraints are the reason this potential was not used before. The Belgian policy represents an easily implementable approach to using this potential without obvious negative effects.

This paper documents how the inequality of household equivalent disposable income has changed in Sweden over the period 1990–2021. We find that income inequality has unambiguously increased. Measured by the Gini coefficient, inequality increased from around 0.19 to almost 0.3 by the end of 2020. We then analyse the backgrounds to this change by measuring the importance of changes in different components of the overall income distribution: the wage distribution; the distribution of hours of work; capital incomes; income differences between labour market participants and non-participants; income redistribution through income taxes and benefits; and, finally, the effect of increased immigration to Sweden.

Children hinder parents' labour market attainments and are often thought to be the reason for the gender pay gap, though its existence is disputed in the long-run. We investigate the importance of using a genetically similar control group for child penalty estimations using detailed Swedish administrative data coupled with the Swedish Twin Registry. We apply an ACE-model to decompose the variance in child penalties among twins into a genetic and common environment part and employ relative fixed-effects regressions to quantify the child penalty for different subsamples of relatives. Using our preferred sample of genetically identical twins, we find a large child penalty of 60.9% but no evidence for a long-run child penalty.