Direct and spillover effects of board gender quotas: Revisiting the Norwegian experience