The unemployment insurance reform introduced in the Netherlands on 1st January 2004 increased the labour supply of individuals above 57.5 years old - age when individuals are likely to be grandparents (generation 1). We investigate the spillover effect of this reform on their grandchildren (generation 3). We implement a IV strategy which exploits variation yield by the policy introduction date and the age eligibility criteria to instrument for grandparents being active in the labour market. We find a positive impact of grandfathers' activity on the grandchildren's educational outcomes. We provide evidence on potential direct mechanisms driving this relationship - i.e. going directly from grandparents to grandchildren - as well as potential indirect mechanisms - i.e. mediated by changes in mothers' (generation 2) labour supply and fertility.